Colonial Williamsburg Digital Library

Research Design

In this study, archaeological and historical documentation are used as complementary sets of data. It is in this integration of the documentary and archaeological evidence where lies one of the major strengths of historical archaeology. Through the use of documents, archaeologists progress beyond inferences based on the archaeological data alone (Spencer-Wood 1987:2; Crumley 1974; Leone and Crosby 1987; Beaudry 1984), with both sets of data combining to provide a more detailed picture of eighteenth-century Williamsburg and its inhabitants than either could give separately (Deetz 1977; Noìl Hume 1978:27; Spencer-Wood 1987:2). Several recent studies effectively combine historical and archaeological data to make statements about social and economic change (e.g., Felton and Schulz 1983; Herman 1984).

An important advance in the field of historical archaeology has been its recent focus on household analysis (Beaudry 1984; Cressey 1983; Kramer 1982; Mrozowski 1981, 1984). A household has been defined as a "co-resident domestic group" composed of "those who share the same physical space for the purposes of eating, sleeping, taking rest and leisure, growing up, childrearing, and procreating"(Laslett 1972:24, as quoted in Beaudry 1984).

The household, which serves as the basic social unit for most cultures, has long been used by as an analytical unit by anthropologists (Goldschmidt and Kunkel 1971; Horne 1982), historians (Lemon 1972, 1980; Henretta 1978; Greven 1966; Demos 1970; Macfarlane 1977), and prehistoric archaeologists (Longacre 1970; Reid and Whittlesey 1982; Netting 1982; Hill 1970). Household archaeological assemblages have been shown to contain the greatest potential for studying changes in social and economic stratification (Cressey 1983:9; Spencer-Wood 1987:9), and techniques have been adapted 5 to isolate those assemblages (Mrozowski 1984). Recent work has successfully made use of household analysis (Baugher and Venables 1987; Dyson 1982; McBride and McBride 1987; Mrozowski 1984; Shepard 1987; Stachiw 1989) to discover meaningful patterns of behavior.

Since research has shown households to be a basic unit of analysis, it is imperative that archaeologists be able to isolate household data from their sites. This has been accomplished successfully on urban lots through the combination of archaeological and historical data (Moran et al. 1982; Mrozowski 1984). Fortunately, excellent documentation provides an almost unbroken chain of title for the Brush-Everard property. Artifact assemblages chosen for analysis were primarily from discrete features which had been used for the deposition of garbage, such as abandoned privies, trash pits, or gullies. All features chosen for analysis had a tightly datable range of deposition, usually less than twenty years. Additionally, horizontal layers of occupation debris, known as sheet refuse, which had been created through activity on the lots were excavated and used for analysis. Although more difficult to attribute to household occupations, those layers which could reasonably be assigned to a household through ceramic dating and crossmend information were also used. Periods of property ownership were compared with each assemblage's ceramic date ranges, using a technique discussed in Mrozowski (1984). Other specialized artifact evidence, such as inclusions of gunsmithing or pharmaceutical debris in several of the analyzed features, also indicated specific household associations.

This archaeological study of the Brush-Everard site focuses not only on the function of the property, but also examines socioeconomic data. Through the analysis of glass and ceramics in archaeological assemblages at the household level, information on social and economic changes occurring at Brush-Everard are determined. Households, it should be stressed, are not strictly tied to kinship with the property owner, but include others living on the property at that time, such as indentured servants, lodgers, and slaves.

While attributing archaeological assemblages to households proved to be fairly simple, the task of isolating African-American slave material possessions from those of their owners was much more difficult. It had been hoped that the two northernmost structures appearing on the Frenchman's Map would be discovered through archaeology and would reveal concentrations of domestic material that could be related to a slave occupation. Unfortunately, all traces of both buildings had been destroyed by more recent construction activities in this area. A line of structural postholes located east of the reconstructed dairy may indicate a possible slave house, but modern disturbance of the soil in the area east of the building made attributing a function and occupation date tentative at best. All of the garbage from the Everard household, from black or white occupants, appears to have been dumped together in a large gully at the north end of the property. Separating slave food and material remains from those of Everard, his daughters, and guests, proved impossible. Numerous potential "markers" or artifacts believed to have been associated with slaves on the property were discovered, however, and are analyzed in a later chapter of this report.

Historical Background

Like most properties located within the confines of the colonial town limits, the Brush-Everard lots have experienced a long and varied history. Following will be a summary of the property history, with more complete descriptions of individual households provided in later. Additional information can be found in the Brush-Everard property history, prepared in 1956 by Mary Stephenson (Stephenson 1956), and in the York County files stored at Colonial Williamsburg's Department of Historical Research.

By the time gunsmith John Brush purchased Lots 165 and 166 in 1717, Williamsburg had been the capital of the Virginia Colony for almost two decades. Growth of the city had been slow during the opening years of the eighteenth century, but increased considerably during the second decade, with lot sales increasing fivefold compared with the period 1700-1710 (Hellier 1989:28). Brush was part of this initial push to purchase property and, in compliance with the Building Act of 1705, constructed a frame house fronting Palace Green within two years of acquisition of the land. When Brush bought his lots, the Governor's Palace, begun in 1706, was virtually complete at the far end of Palace Green. The construction of this building made the land on either side of the green highly prized, and soon other lots along Palace Green began to sell.

Williamsburg, as the new capital, had replaced Jamestown as the Virginia Colony's center of commercial, government and social activities. During this period, the lots on 8 Figure 4. Location of Brush-Everard property on ca. 1800 "College" or Bucktrout map. Block 29 (as the block bounded by Palace Green, Nicholson Street, North England Street, and Scotland Street was defined by Colonial Williamsburg in the early days of the Restoration) comprised a mixed neighborhood of commercial, domestic and industrial properties. There, John Brush plied his gunsmithing trade adjacent to the home of the Governor, while William Levingston constructed a theatre and a bowling green on his nearby lots. Levingston's home also functioned as an ordinary, no doubt serving visitors who came into Williamsburg on business.

Brush, a gunsmith, retained Lots 165 and 166 until his death in 1727, at which time the property passed into the hands of his unmarried daughter, Elizabeth Brush, and his son-in-law, Thomas Barbar (husband of his daughter Susannah). Elizabeth 9 Figure 5. Brush-Everard kitchen and smokehouse, looking south. Brush subsequently sold her share of the property to Thomas Barbar in 1727. When Barbar died several months later, his widow sold the lots in 1728 to Mrs. Elizabeth Russell (perhaps her now-married sister). Most likely, Elizabeth Russell later married Henry Cary II, because in 1742, Cary and his wife, Elizabeth, conveyed Lots 165 and 166 to William Dering. Dering, a dancing teacher and an artist, experienced financial difficulties and mortgaged the property in 1744 to secure a debt. With the property still mortgaged, Dering apparently moved to Charleston in late 1749. There is no clear title to the lots until 1779, although they may have been purchased in 1751 at an "outcry" by either Thomas Everard or John Blair. Archaeological evidence suggests that Williamsburg Mayor Thomas Everard lived on lots 165 and 166 as early as 1756, when he sold his house and property on Nicholson Street to Anthony Hay (Frank 1967). By 1779, deed books show Everard as the owner of Lots 165, 166 and 172.

By 1750 or so, the character of the lots along Palace Green had changed considerably from that of thirty years earlier. As other towns formed in the western part of the colony, Williamsburg began to diminish in social and economic importance. As the need for consumer services decreased, the commercial character of the Palace Green lots virtually disappeared, and they became almost entirely residential after the early 1760s. With this abandonment of commercial establishments, Palace Green took on the character of a gentry neighborhood, a character which it was to maintain throughout the years preceding the Revolutionary War.

The ownership history of the lots after the Revolutionary War is mostly unremarkable. After Everard's death in 1781, John Stith was taxed for the three lots (165, 166, and 172). Dr. Isaac Hall owned the property 10 between 1787 and 1788, when he conveyed it to Dr. James Carter. After 1798, the property was taxed as part of James Carter's estate. Land tax records between 1820 and 1830 charge the lots and buildings to Milner Peters. From 1830 to 1847, Dabney Browne owned the property, which he transferred to Daniel Curtis in 1847. Curtis conveyed this property to Sydney Smith in 1849. The property remained in the Smith family until W. A. R. Goodwin purchased it in 1928 for the Williamsburg Restoration. The Brush-Everard house was the setting for an early twentieth-century historical novel and became popularly known as the Audrey House after its publication. Between 1949 and 1952, the Colonial Williamsburg Foundation restored the Brush-Everard House and its outbuildings, which today are used as exhibition buildings interpreted to the Thomas Everard period of ownership.

1947 Excavation

During the winter of 1946/47, extensive archaeological work was conducted on the Brush-Everard property under the direction of James M. Knight of Colonial Williams-burg's Architecture Department. The excavation was divided into three areas (Fig. 6), as shown on the revised 1932 Archaeological and Research Key Map (Perry, Shaw & Hepburn 1932). Each of these areas has a designation composed of its block number, which in this instance is Block 29, and a letter indicating its so-called (and arbitrarily defined) "archaeological area." Each time archaeology was conducted within a block, a letter designation was assigned to the area excavated; for instance, the first area examined on Block 29 would be given the designation 29A, the second section 29B, and so forth. The three areas examined in 1947 were assigned the designations 29D, 29E, and 29F.

Colonial Williamsburg's standard archaeological procedure prior to 1957, when English archaeologist Ivor Noìl Hume arrived to take over the program, involved the use of cross-trenching as an investigative technique. Placing narrow trenches at 45 degree angles to the property lines, spaced five feet apart and dug to the depth of sterile subsoil, would generally guarantee that any structures containing brick foundations would be located. Since the purposes of the early excavations were to locate buildings while working under tight time and budgetary constraints, this technique as a rule worked very well. Fig. 7 shows examples of this cross-trenching technique as used on the Brush-Everard property during the 1947 excavation.

Historical sources suggested that the cross-trenching technique would be particularly fruitful on the Brush-Everard property. The Frenchman's Map of 1781 showed six structures standing on the property during the Revolutionary War. Of these, only two, the house and the kitchen, remained standing 12 Figure 6. Archaeological areas cross-trenched in the 1940s. until the present day. Since the purpose of the Frenchman's Map was to show buildings which could be used for billeting soldiers, smaller structures such as smoke-houses and dairies were generally not depicted on the map. The remaining four buildings shown appeared to be at least as large as the kitchen.

James Knight and his crew located the remains of a laundry, a dairy, a well, a smokehouse and an office during their trenching, and with the reconstruction of these buildings the property was restored to what was then believed to have been its late eighteenth-century appearance. Their efforts failed, however, to locate traces of two (perhaps three) of the outbuildings portrayed on the Frenchman's Map in the northern part of the property. A number of factors could have been involved in the failure to find these buildings, including destruction by later landscaping and construction activities, or more likely, that the buildings were supported on wooden posts or brick piers. The remains of buildings with these types of structural foundations were generally not found using archaeological cross-trenching.

James Knight prepared a summary archaeological report of his investigations, as well as a detailed map of the brickwork that was located (Knight 1947a, 1947b). Both of these have been used to describe the results of his work. Discussion of these findings will be grouped according to his archaeological areas.

1967 Excavations

The 1967 excavations around the Brush-Everard kitchen were conducted using excavation techniques first introduced in the 18 Historic Area by Ivor Noìl Hume in the late 1950s and early 1960s. Stratigraphically controlled digging provides archaeologists with the ability to chart changing conditions within narrow time ranges on properties, and to make statements about specific inhabitants, their activities, and their social and economic status. The 1967 excavations, while providing information about the original kitchen, also furnished details about eighteenth-century property divisions, and how the property was used by its various residents.

Under the direction of Archaeologist Ivor Noìl Hume, an in-depth excavation of the Brush-Everard kitchen and surrounding area was performed in early 1967. The area covered by these investigations is shown in Fig. 12.

Results of Archaeological Monitoring, 1982-1987

Since the formation of the Department of Archaeological Research in 1982, four projects which necessitated archaeological monitoring have taken place on the Brush-Everard lot. Each of these monitoring episodes was instigated when a utility line was installed or repaired. A member of the Department of Archaeological Research was present during these projects, in order to protect archaeological resources, record archaeological stratigraphy, and recover artifactual material. At best, this monitoring is a last-ditch effort to record those areas being destroyed in the wake of human progress, but it can also serve a useful function in allowing archaeologists a "free" (of sorts) look at areas they may never otherwise get a chance to examine. Each of these monitoring activities will be discussed below, along with the effect they had on guiding the course of the 1987 excavation.

Summary of Previous Excavations of Brush-Everard

Although what seems like a great deal of archaeological research has been conducted on the Brush-Everard property prior to 1987, in actuality, the only areas which have been exhaustively studied are those excavated during the 1967 investigation. Knight's cross-trenching proved, fairly conclusively, that no structures with continuous brick foundations were located in the area between the laundry and the Palace Power Plant. The potential for post- or pier-supported buildings still remained a strong possibility, however. Monitoring of utility trenches in the area showed a great deal of twentieth-century disturbance, so chances of finding the map-predicted buildings could have been compromised by later activity. The ash and rubble layer beneath the redeposited yellow clay in the northwestern portion of the site, however, suggested the potential for this area containing structural debris of a building or buildings constructed, occupied, and destroyed prior to the rendering of the Frenchman's Map in 1781/2.

Field Methods

The excavations between 1987 and 1989 were conducted on Lots 165 and 166 of the Brush Everard property, north and south of the house. Since map and documentary evidence, as well as the results of utility trench monitoring, had suggested the probable presence of eighteenth-century structures north of the house, a strategy of "open-area excavation" was deemed the best technique for locating these buildings. Excavation areas were located in places shown by maps and remote sensing to be those most likely to yield structural remains (Fig. 16).

Figure 16. Areas excavated 1987-1989.

Prior to beginning the excavation, a grid system was imposed over the site for horizontal control of archaeological features and objects. A baseline was established, running north-south, parallel with Spotswood Street, within the Brush-Everard property. An arbitrary datum for the work was an iron rebar set at point 200N/80E. From this point, a system of 10-foot squares was placed over the entire area encompassed by Lot 166. Vertical control over the site was established by using a dumpy level to measure the position of soil strata and archaeological features in relation to sea level. The level was set up each day over the rebar at 200N/80E, and elevations were tied in with a known benchmark.

Color and texture changes in the soil were used to guide the excavation of the site. Color was determined by removing a small 24 chunk of undisturbed moist soil and comparing it with Munsell color charts (Munsell 1975). The main soil color was noted, as well as any mottlings or streakings. Soil texture was determined through analyzing the percentages of sand, silt, and clay contained within the soil and comparing this with a soil chart published in Susan Limbrey's Soil Science and Archaeology (Fig. 17). Any inclusions, such as brick, shell, mortar and charcoal were also noted.

With some exceptions, mason's trowels were used to excavate all undisturbed soil layers and features. Only topsoil, the backfill of the 1947 archaeological trenches, and the fill of utility trenches were removed using shovels. All soil, except that noted above, was screened through ¼" wire mesh, to facilitate the recovery of small finds. Cultural material from each layer was retained, with the exception of brick and charcoal, whose presence was simply noted. Information about soil color, texture, inclusions and stratigraphic position was recorded on a context record form. All layers and features were mapped in plan and features were bisected and mapped in profile after half their fill had been removed. Photographs of archaeological features and strata were taken as necessary.

Figure 17. Soil triangle (after Limbrey 1975:261).

Archaeological sites are made up of two components: features and layers. These components occur with varying frequency, but tend to increase in number with the length of site occupation. At a certain point the density of features and layers, especially on urban sites, causes site interpretation to become increasingly difficult. One traditional way to simplify interpretation is to create an overall "composite" map of all relevant features and layers, color coding each feature and layer based on artifact dating and stratigraphic sequences of these components. This system of interpretation works well when the density of features and layers is not great, thus making possible to ascertain any patterns that represent the remains of fencelines, post-set buildings, formal gardens, and activity areas. As the density of activities becomes greater, however, it becomes increasingly difficult to ascertain such patterns. The urban nature and long occupation range of the sites inside Colonial Williamsburg's Historic Area result in such a high density of activities through time that the use of composite maps as an interpretive tool is greatly hampered.

When confronted with a similar problem, though of even greater magnitude, Bermudian (formerly English) archaeologist Edward Harris developed the so-called "Harris Matrix" approach. The Harris Matrix itself is a flow chart that illustrates stratigraphic relationships between layers and features. Layers can be formed either by nature (erosion, for example) or by humans, and consist of thin deposits of soil spread over a large area. These layers can develop over a relatively long period of time and are usually the product of one particular activity. When an activity ceases or changes, the layer formation process ends and a new layer starts to form. Subtle changes in the landscape can cause a particular layer be replaced by a different soil formation. An 25 example of this would be the construction of a fence, which may be related to changes in patterns of land use and therefore cause an alteration of soil formation. Possible changes in behavior such as pedestrian traffic flow patterns, or an altering of trash disposal would also result in changing layer formation processes.

Archaeological stratification may also be formed by excavation of holes or pits (features) by man or animals. Features are generally, although not always, dug and filled in a short period of time and can vary in size, shape, depth and composition. Some features are filled with soil all at once, while others are left open and filled through a slow silting activity. When this occurs, the feature can be filled with various different soil layers that accumulate through time. A feature actually has two components: its fill, the actual soil used to fill in the concavity, and its cut or (as Harris calls it) its "interface". An interface is a rather abstract concept that represents the actual act of digging the hole. For a more detailed description of interfaces and their importance to the study of stratigraphy, see Harris' book Principles of Archaeological Stratigraphy (Harris 1979,1989).

Each layer or feature was assigned a sequential "Harris" number that represented the entire layer or feature. Each number was added to the flow chart under the numeric designation of the layer that sealed or was on top of it. All features that intruded a particular layer were grouped together and the layer that they intruded was added beneath the features. All features were assigned an interface number regardless of the type of filling episode and the interface number was added to the chart under the last fill number of each feature.

The actual charting of the stratigraphy at Brush-Everard was fairly simple even with the large amount of disturbance present, since the soil layers were fairly uniform and extended over large areas of the site. A recurring problem in the Historic Area has been that layers are often similar in color and composition, making separation of layers very difficult. A system was initiated to insure that layers created slowly over long time periods were not accidentally lumped together, obscuring any data manipulation involving these layers. Any layer that was over 0.25 feet thick was subdivided into separate layers. After the artifact inventory was completed, deposition dates of the subdivided layers were compared and a determination was made to lump the layers because of similar dates or have them remain separate for analytical purposes.

When the Harris Matrix flow chart was finished it was easy to understand, not only the sequence of layers, but the time relationship between features . Instead of a map showing all features excavated on a site, the Harris flow chart allow the production of phase plans that recreate what the site looked like during a particular period. Each layer of interest receive its own plan with all features that existed directly after the layer was formed. Analysis is simplified because only features that date to a particular time period are represented and the problem of high density of feature activity is resolved.

The recovery of microfauna, botanical materials and oyster shell from the Brush-Everard site required specialized sampling procedures. All features, excluding postholes, postmolds and utility trenches, were screened through both one-quarter and one-eighth inch stacked wire mesh screens. Small bones and other artifacts remaining in the one-quarter-inch screen were carefully removed and bagged. Any soil and material that did not readily fall through the one-eighth-inch screen was placed into marked plastic bags for later waterscreening. After waterscreening, the remaining material 26 was sorted according to category (seeds, small bone, eggshell, crab claws, beads, ceramics, straight pins, etc.) and added to the artifact inventory. In order to sample botanical material, ten-liter samples of soil from each layer and pertinent feature were bagged without screening. All oyster shell (with the exception of small fragments) were retained from soil layers and features, excluding postholes, postmolds and utility trenches.

Laboratory Methods

Processing of the Brush-Everard artifacts-washing, numbering, inventorying, and cross-mending-was conducted in the laboratory of the Dept. of Archaeological Research. Initially, incoming groups of artifacts were received on a daily basis. At this stage, a laboratory technician recorded the context(s) represented, the number of bags within each context, and the date each was received.

All logged-in artifacts were washed, with extremely delicate or unstable artifacts removed for conservation during this phase. Following cleaning, several categories of artifacts, including shell, brick, and coal, were simply counted and re-bagged. Animal bones from each context were counted and removed to the zooarchaeology laboratory for further processing and analysis. All other artifacts were retained.

After washing and sorting, artifacts were labelled with their provenience number so that context information would not be lost during analysis. Artifacts were inventoried at this time, using a micro-computer-assisted system developed by the D.A.R. Numeric codes were assigned for specific descriptive information, in order to permit easier data manipulation. The type, decorative attribute(s), technology, mark(s), and other distinctive characteristics were recorded for every artifact.

At this time a terminus post quem (TPQ) date was given to each unit. The terminus post quem represents the earliest date that the context could have been deposited, and it was assigned by determining which artifact had the most recent initial documented date of production.

Cross-mending was the final phase of processing, and was conducted only on selected objects. Its purpose was to demonstrate relationships between various layers and features, through showing that portions of the same ceramic or glass object were deposited in each. Two categories of relationship were noted: cross-mends (glue-fits between sherds from different contexts) and non-contiguous relationships (sherds that do not glue together but are most likely from the same vessel). All relationships, whether cross-mend or non-contiguous, were recorded. Cross-mending was conducted on ceramics, wine bottle glass, and table glass from eighteenth-century contexts.

Another important purpose of cross-mending was to reveal the minimum number of vessels or objects present on the site. A "vessel" or "object" was determined to be one or more fragments that represented what had most likely once been a single entity. Each vessel or object was assigned a unique vessel (U.V.) number, and each sherd of every vessel was labelled with its corresponding U.V. number, so that cross-mend information would not be lost.

Cross-mending of ceramics, wine bottle glass, and table glass produced a minimum vessel count. Minimum numbers of vessels were determined using either rims or bases-whichever was in greater abundance-as the initial count. Distinctive vessels were added subjectively by the lab staff, based upon their unique decorative or technological 27 attributes. Vessel counts were compiled conservatively so as to provide an actual minimum number of vessels. This is not to say that there were no additional vessels present-there undoubtedly were-but this was the least number represented in the assemblages.

Minimum vessel information was grouped into four household groups: the John Brush household (1717-1727), the George Gilmer household (1735-1757), the Thomas Everard household (c. 1750-1781), and the Polly Valentine household (c. 1840-1865). This information was additionally quantified by ware type, vessel form, and vessel function.

Artifact records are stored in the department laboratory. Selected artifacts, based on their representativeness, uniqueness, or potential for further research, were catalogued, assigned an individual catalog number, and housed in drawer storage in the D.A.R. study collection. All other artifacts are stored in the Archaeology Warehouse.

Borrow Pit
(Harris Numbers 231, 232)

One feature, north of the present house on Lot 166, consisted of a large circular hole, approximately 15 feet in diameter, excavated 2.5 feet into the clay subsoil (Fig. 18). This pit was left open to partially fill with silt, and was later deliberately filled with yellow clay. Very few artifacts were found in the pit fill; these included only non-datable materials, such as animal bone and pipestems. Two test trenches, one running north-south and another running east-west, were excavated through the feature. In total, approximately one-third of the feature was excavated.

Figure 18. Location of clay borrow pit.

At the present time, it is believed that the pit may have been excavated during the seventeenth century as a borrow pit for the recovery of brickmaking clay. The hole was left open, silting in to a depth of about eight inches, until it was later filled with clay, possibly when John Brush excavated the cellar for his house shortly after purchasing the lot in 1717. The stratigraphic position of this pit, sealed by an early eighteenth-century layer and cutting through sterile soil layers, suggests the seventeenth-century date assigned for the creation of the feature. Borrow pits, sometimes dug near the site of seventeenth-century dwellings, were often used as garbage dumping areas for the residents of these structures (Edwards 1987). The lack of artifacts recovered from the fill of this pit suggests that the structure for which these bricks were intended did not stand nearby.

Ditch and Gully
(Macro-Feature 14 Harris Numbers 91 and 102 ER1256R and ER1261L)

A second Middle Plantation period feature, consisting of two parallel trenches, was discovered on the southern portion of Lot 165, south of the reconstructed garden shed (Fig. 19). Portions of this feature, running southwest to northeast, had been discovered extending under the kitchen and smokehouse during the 1967 excavation. At that time, these trenches were interpreted as a square-sided ditch (ER1261L) and a natural gully (ER1256R).

The square-sided ditch (Harris Number 91 and 102), 1.5 feet wide, cut to a depth of 1.5 feet into the sterile clay subsoil. Although a 40 foot segment of this ditch was excavated in 1967 (ER 1261L), no artifacts were Figure 19. Plan and profile of palisade trench. 31 recovered from its mixed yellow clay fill, precluding any precise dating of the feature. Since this trench ran under the foundation of the 1730 kitchen, however, it can be safely dated as having been created and filled before the kitchen was constructed.

A larger 4-foot-wide ditch (ER 1256R) ran parallel with the square-sided trench approximately 8 feet to the south. Filled with a hard grey sandy loam, the ditch contained sloping sides, a rounded bottom and was a little over 1 foot in depth. A 20-foot portion of the ditch was excavated in 1967. There were no artifacts in the fill of the ditch, suggesting that it too had been completely filled before the development of the property in the second decade of the eighteenth century.

During the 1988 excavation, additional portions of the square-sided ditch were uncovered and excavated. This feature appeared as a 1.2-2.0 foot wide trench cut 1.5 feet deep through sterile clay subsoil, sealed by a layer of soil whose deposition dated to the early eighteenth century. Portions of the trench excavated in 1988 contained two distinct types of fill, leading to its reinterpretation as a slot fence. A dark brown (Munsell color 10YR4/3) sandy loam fill, visible along one edge of the trench, is believed to have been created by the decay of wooden fence posts placed upright against one side of the ditch. The second fill, a mottled brownish yellow (10YR6/8) sandy clay with sparse oyster shell and charcoal flecking had been packed against the wooden posts to hold them in place. No artifacts were recovered from either of these two fills to assist with further dating of the trench.

Similar trenches identified as slot fences have been discovered on various seventeenth-century sites, including Wolstenholme Towne (Noìl Hume 1979), Nansemond Towne in Suffolk (Luccketti 1988) and along the Hampton River (Edwards et al. 1989). These fences were constructed by first placing planks upright side by side in a narrow, fairly shallow ditch with soil packed around the boards to secure them in place. Such slot fences were fairly insubstantial and used primarily to keep animals from gardens or yards (Edwards et al. 1989:92).

Archaeological Results

A number of excavated soil layers and features related to the John Brush occupation (1717-1727). These included a sheet refuse layer, a privy, a trash pit and two layers of debris in a ravine on the northern extent of the property. Each assemblage will be discussed here in terms of its composition, artifacts, and any specialized analysis that was completed.

Primary Ravine Layers
(Macro-Feature 1 29F-Harris 140, 154, 141, 155, 166, 333, 334)

The earliest soil layer in the ravine consisted of a dark brown (10YR4/2) silty clay loam, 41

Table 1.
Ceramic and Glass Vessels from the Brush Trash Pit

Type	Form	No. of Vessels	% of Total
Chinese porcelain	Cup	2	13.3
Dipped white salt glazed stoneware	Hollow	2	13.3
	Mug	1	6.6
Yorktown-type	Milkpan	1	6.6
	Chamberpot	1	6.6
Glass	Wine bottle	5	33.0
	Pharmaceutical bottle	1	6.6
Leaded glass	Wine stem	1	6.6
Ceramic fireclay	Crucible	1	6.6
Total		15	100.0

Note: One of the two dipped white salt glazed hollowares crossmends with sheet refuse Layer 153/103. Figure 28. Profile of John Brush-period ravine layers. containing charcoal and brick flecking. This ashy layer, which ranged from 0.3-0.4 feet in thickness, sealed sterile subsoil and can be linked to the earliest years of occupation (c. 1717-early 1720s) on the property. A range of domestic and architectural debris dating to the first quarter of the eighteenth century were recovered from this layer (Fig. 29).

Twenty-three ceramic vessels were recovered from the earliest ravine layers (Table 2). Fragments of three delft fireplace tiles decorated in blue suggest a delft fireplace surround, and leaded window cames from the ravine and the trash pit excavated in 1967 indicate casement windows were in use somewhere on the property, perhaps in the first period house or shop. No identifiable gun parts, and no gunsmith-related material other than slag and coal, were recovered from the primary ravine fill, further corroborating the idea that Brush's shop was located at the opposite, or southern, end of the property, where numerous gun parts were recovered.

42

Table 2.
Brush Ravine Layer-Primary Ash Layer
Vessels Organized by Form

Type	Form	No. of Vessels	% of Total
Yorktown-type	Chamberpot	1	4.3
	Bowl	1	4.3
	Milk pan	3	13.0
Delftware	Fireplace tile	3	13.0
	Cup	1	4.3
	Plate	3	13.0
	Dish	1	4.3
Westerwald stoneware	Chamber pot	1	4.3
Nottingham stoneware	Bowl	1	4.3
Fulham stoneware	Storage jar	2	8.7
White salt-glazed	Milk jug	1	4.3
Chinese porcelain	Plate	1	4.3
	Saucer	1	4.3
Fulham stoneware	Mug	1	4.3
Red sandy ware	Milkpan	1	4.3
English stoneware	Jug	1	4.3
Total		23	100.0

Feature Description

This building was first discovered during Noìl Hume's 1967 excavation, when the intact southwest corner of its brick foundation was uncovered and mapped. Although time constraints prevented the excavation from proceeding any further, the report concluded that this was the robbed wall line of a colonial period building.

Figure 33. Plan on privy foundations.

The primary goal of the 1988 excavation along the southern border of the Brush-Everard property was to uncover the remains of a nineteenth-century slave house. Once mains of an 8 x 10 foot structure, which had once comprised a frame building seated on a brick foundation. Intact brick remained only at the northeastern and southwestern corners of the building, the rest of the brick having been removed, probably for reuse elsewhere. Traces of the foundation were evident in concentrations of crushed brick which were present extending south and west from the northwest corner of the structure. The interior of the building contained a number of clear soil layers extending into a deep pit which had been dug inside the building.

The pit, which measured 5 x 6 feet, was originally excavated to a depth of five feet into the surrounding subsoil clay. Containing slightly sloping walls and a flat bottom, 50 Figure 34. Profile of privy pit. the pit showed no evidence that it had been lined with brick or wood. A water line, which had been installed to service the Brush-Everard House sometime early in the twentieth century, cut through the center of the privy foundation and pit, causing some soil disturbance down to a depth of 2.5 feet below the present ground surface. Although the placement of this water pipe made excavation logistically difficult, it did not significantly damage the integrity of the feature.

Excavation Techniques

After the structure was revealed in plan view through removing the overlying soil layers, the building was divided into four equal-sized quadrants for excavation. This was done to gain some control over artifact distribution within the feature. Each layer or tip of fill was assigned a layer number and a suffix representing the four subdivisions within each layer (a-d). The feature was hand excavated and the soil screened through one-quarter and one-eight-inch wire mesh. Ten liter soil samples were saved from each layer of fill for flotation, seed, pollen, parasite and phytolith analysis. All artifacts were retained, with the exception of brick, mortar and charcoal, of which samples were taken.

The Privy as an Exercise in Site Formation Processes

The archaeological record is created through sets of events which have been labeled "formation processes" (Schiffer 1987). Schiffer discusses two main types of formation processes; cultural formation processes, which are the result of human behavior, and non-cultural formation processes, resulting from environmental events such as erosion, decay or seimic activity.

Through examining the privy remains, the composition of its soil layers, its artifacts and ecofacts, and the relationships of these 51

The soil was examined in terms of its physical composition, its chemical components, its placement within the privy pit, and its artifactual content. Each of these factors was used to help determine the origin of the soil layer and what information it could provide about the life cycle of this structure, and on a more general level, about the Brush-Everard property as a whole. The artifacts were categorized by type and examined for patterns. It appears that the soil layers were generated by three distinct activities: (1) actual use of the privy, (2) silting or erosion, and (3) deliberate deposition of household, construction and industrial debris (Table 5).

John Brush purchased his lots on Palace Green in 1717 and, in accordance with city building ordinances, had completed his house within two years. The construction of the privy probably took place during the house construction or soon thereafter. The privy was located along the southern perimeter of the property, at a distance of 20 feet from the house. Sometime during the ten years Brush owned the property, he removed the frame superstructure from the building foundation and began to fill the privy pit with gunsmithing and household debris. The pit appears to have been filled quickly, probably within less than a year, with any remaining brickwork dismantled for use elsewhere. One possible explanation for the rapid demise of this building may be the fact that it straddled the property line between Lot 165, owned by Brush, and Lot 164, owned by ordinary-keeper William Levingston. Although no evidence of a boundary dispute has been located in the York County records, it is possible that Levingston, or later owner Archibald Blair, who mortgaged the adjoining property from the financially troubled Levingston in 1721, may have wanted the building removed.

Of the sixteen layers of fill within the privy pit, only the bottom five (Layers 169-173) showed evidence that they were generated by actual use of the privy. These layers showed mixing, as might be expected in the wet environment typical of a functioning privy. Interspersed within several of the layers were washed lenses of silt. The bottom three layers were grey in color and, although 52 not particularly organic in appearance, contained dietary pollen, vegetable and fruit seeds, and human intestinal parasites. Additionally, soil testing showed that these layers contained very high levels of phosphorus. Analysis at the King's Reach site (Pogue 1988) and at Clocker's Fancy (Keeler 1978), both colonial sites in Maryland, showed high concentrations of phosphorus in privies, animal pens, and in areas where chamberpots had been emptied. These bottom layers also contained proportionally fewer artifacts than layers deposited later, in what is believed to have been deliberate disposal of household and industrial debris.

Interestingly, of the eleven complete or completely mendable ceramic and glass items discovered in the privy pit (Fig. 35 and 36), ten of these were recovered from layers believed to have been created primarily through human waste disposal (Fig. 37). Additionally, six white clay tobacco pipes, with unbroken bowls and at least six inches of stem attached were also found in the bottom layers of the privy (Fig. 38). Recovery of unbroken or virtually reconstructable items from archaeological contexts is rare, and usually indicates some specialized discard behavior or loss. The pipes, ceramics and glass items may perhaps most reasonably be explained as the result of loss during activities such as smoking and drinking in the privy building.

Figure 35. Delft teacup and teabowl recovered from the Brush privy.

Figure 36. Complete wine bottles from privy.

Some 99% of the material from Layer 169 was animal bone. Some 543 bones were found, including, in addition to the nearly-complete skeletons of at least two cats, dietary remains of pig, cow, striped bass, chicken and sheep or goat.

Layers 167 and 168 appear to have been created through erosional silting after the privy superstructure was removed from its foundation. Composed of silted sand particles containing small patches of yellow subsoil clay, these layers were found only along the perimeter of the pit. As might be expected, these layers contained low artifact totals (19 and 86 items, respectively).

The remainder of the soil (Layers 139-166) appears to have been created as the result of deliberate pit filling. The artifacts, sizable in number, suggest three sources for the garbage: household trash in the form of broken ceramics, glass and faunal bone; waste and gun parts from Brush's gunsmithing operations; and architectural debris from a renovation or building episode, probably the renovation of the main house.

Artifacts

Excavation of soil layers from the interior of the privy building and its pit revealed a total 53 Figure 37. Profile of privy showing layers containing unbroken artifacts (Note: For a complete profile of privy with layer descriptions, see Fig. 34) of almost 5,000 artifacts. These artifacts included the typical range of household items recovered through excavation on Anglo-American sites of the early eighteenth century (Fig. 39). Aiding in attributing this feature to the John Brush household were numerous artifacts which could be associated with Brush's gunsmithing operation, namely, gun parts, gun flints, gunmaking tools, and industrial waste in the form of slag and coal. The smithing debris was scattered throughout the privy (with the exception of the bottom five layers), strongly suggesting that the feature was completely filled during Brush's ownership of the lot (Fig. 40).

Faunal Analysis

Some 2337 bones from the Brush privy were analyzed in 1990-91 by Stephen Atkins of Colonial Williams-burg's Zooarchaeology Lab. At least twenty-one species were represented (Table 7), including a variety of fish, turtles, domestic fowl, wild mammals, and commensal species such as cats and rats. The overwhelming majority of remains, however, both from the standpoint of raw numbers and even more so from the standpoint of meat contributed to the diet, came from the major domestic livestock-cattle, swine, and sheep or goats.

Biomass estimates (described in Appendix 3) indicate that, like virtually every other site in the Chesapeake, this one was dominated by cattle, with 69.6% of the utilized meat (Fig. 43). Pig was a distant second at 11.5% of the total biomass. No other taxon represented more than 1.7%. It seems clear that beef was by far the predominant meat eaten; as Bowen (1992) has noted, it appears to have been much later that the impression arose that it was in fact pork, rather than beef, that was the signature meat of historic Virginia.

Sheep or goat (Ovis aries/Capra hircus) was present in small numbers, as was white-tailed deer (Odocoileus virginianus). Small mammals, including rabbit (Sylvilagus floridanus) and grey squirrel (Sciurus carolinensis), were probably eaten. Horse, cat, and rat were also found, but are unlikely to have been dietary remains.

A variety of birds were found, including duck, goose, chicken, and turkey. All were kept as domesticates in the eighteenth century, although wild ducks and geese were also commonly taken.

A few turtle bones were found, and turtles appear regularly in eighteenth-century receipt books. A few frog or toad bones most likely were from accidental visitors.

A number of commonly-taken fish were found, including sturgeon (Acipenser spp.), yellow perch (Perca flavescens), striped bass (Morone americana), black drum (Pogonias cromis), red drum (Sciaenops ocellatus), and weakfish (Cynoscion spp.).

Oyster Shell Analysis

Oysters (Crassostrea virginica) were an important and abundant food source in colonial Virginia; their shells are found in great numbers on seventeenth- and eighteenth-century Chesapeake sites. Recent advances in the studies of oyster shell allow information about oyster harvesting, marketing, and preparation to be gleaned from this source. An extended discussion on oyster shell analysis and its ramifications can be found in Kent (1988); the specific results from the analysis of the oyster shell from the Brush privy will be discussed here.

Figure 43. Relative dietary importance based on biomass.

The privy oysters were examined individually in three ways. The dorsal-ventral/anterior-posterior dimensions of each shell were first measured in order to determine the height to length ratios (HLR) for each shell (Fig. 44). This measurement can be correlated to marine environment (Galtsoff 1964:18-20). For example, an oyster which grows on a firm, sandy marine bottom, known as a sand oyster, will develop a shell whose HLR is less than 1.3. Bed oysters, which grow on a mixed sand and mud substrate, have an HLR of between 1.3 and 2.0. Channel oysters (which grow in clusters) and reef oysters have an HLR of over 2.0. By analyzing height to length ratios, the marine environment from which the oysters were obtained can be determined.

Secondly, the shells were examined for the presence of attached organisms. Many marine organisms attach themselves to oysters while they are in their natural environment, 58 Figure 44. Height-to-length ratio (HLR) for oyster shells (after Kent 1988). and the presence and distribution of such organisms can assist in interpreting environmental conditions, most particularly, salinity of the water.

Thirdly, the shells were examined for scarring and damage associated with food preparation.

Some 231 shells were recovered from the privy. All complete valves and broken shells with intact hinges were saved for analysis. The shells were first separated into upper and lower valves, which broke down as follows:

Since more lower valves were recovered, analysis was limited to these components.

Measuring height to length ratio shows that the shells were primarily from bed oysters (Fig. 45). This type of oyster would be consistent with that recovered by standard oyster harvesting methods of the early eighteenth century. The Swiss visitor Frances Michel wrote in 1701 of oystering in Virginia:

The abundance of oysters is incredible. There are whole banks of them so that the ship must avoid them.... The inhabitants usually catch them on Saturday. It is not troublesome. A pair of wooded tongs is needed. Below they are wide, tipped with iron. At the time of the ebb they row to the beds and with the long tongs they reach down to the bottom. They pinch them together tightly and then pull or tear up that which has been seized(Wharton 1957:37-38).

Figure 45. Oyster types from John Brush privy.

Hand tongs such as these were first introduced in the Chesapeake Bay area around 1700 (Chowning 1990:87). The Brush oysters, it appears, were not recovered along tidal flats, but were instead tonged from deeper water.

Oysters can tolerate a wide range of water salinity, but the presence, distribution and density of organisms such as boring sponges, polycheate worms, and boring clams indicate general salinity regime from which the oyster originated. Salinity regimes are ranked from I to IV, with I representing the lowest level of salinity. Analysis showed that the shells from the privy were harvested from a range of water salinity levels, suggesting recovery from different regions of the nearby rivers and bay (Fig. 46). This suggests 59 Figure 46. Water salinity ranges of Brush privy oyster shells. that, rather than being collected from a single area nearby by Brush or a servant, an established oyster harvesting and marketing system was in place in Williamsburg by this time and that Brush was probably purchasing his oysters in town from hucksters or at the market.

Food preparation techniques can also be discovered indirectly from oyster shells. Seventy-one percent of the shells which exhibited preparation marks showed signs of hacking. Hacking, or cracking open the shells by hand, would allow oysters to be extracted to be eaten raw or for the preparation of dishes such as scalloped oysters. Oysters were also smoked and dried, fried, pickled and fricasseed in eighteenth-century Virginia (Wharton 1957:46, Jackson 1988:68). Some 21% showed scorch marks consistent with roasting, while only 4% showed shucking scars.

In conclusion, it appears that oysters were an important or preferred food at the Brush household, since the privy contained large numbers of discarded shells. The majority of oysters were bed oysters, recovered by tonging in fairly deep water. The even distribution from all water salinity ranges suggests that an oyster marketing system was in place in Williamsburg by the end of the first quarter of the eighteenth century.

Ethnobotanical Analysis

Excavation of the Brush privy revealed a large sample of botanical material in the form of macrofossils (seeds) and pollen. Analysis of the feature's seeds was undertaken by Dr. Steve Mrozowski of the University of Massachusetts/Boston (see Appendix 4), and pollen and parasite identification was conducted by Dr. Karl Reinhard of the University of Nebraska (see Appendices 5 and 6). Both reports contained information which shed light on the diet and health of the John Brush household.

One of the first tasks undertaken was to determine if the Brush-Everard feature was indeed a privy, as it appeared to be, and if so, which soil layers were deposited in the pit as a result of human waste disposal. Soil samples suspected to contain fecal material based on their physical appearance and position within the feature were submitted for parasite testing, along with a control layer from a garbage deposit from the same site and time period. The privy soils produced numerous egg sacs from two species of parasite: human whipworm (Trichuris trichuria). These species were present in large enough numbers (over 1,200 eggs per gram of soil) to indicate that the soil was indeed derived from latrine deposits (Reinhard 1989). No parasite egg cases were present in the control sample. Testing also revealed a wide diversity of species in the pollen and seed spectra from the same soil strata, a result that also fits the profile typically seen in primary privy fill.

Two main types of pollen were recovered through the testing: wind-pollinated plants, such as trees, grasses and weeds, and insect-pollinated species (Fig. 47). 60 Figure 47. Pollen percentages from the John Brush privy (from Karl J. Reinhard, Texas A & M University) Determining past site vegetation can be problematic using evidence from wind-pollinated plants, since the ease of contamination from wind-borne pollen must be taken into account (Pearsall 1989). Insect-pollinated plants, such as those of fruit trees and vegetables, however, are rarely present in archaeological features unless they have been deposited through human activities, such as food consumption. Pollen grains, along with vegetable and fruit seeds, are ingested with food, and except for their internal cytoplasm, pass through the digestive tract virtually intact. The insect-pollinated species can almost always be attributed to dietary use.

One of the most obvious subjects of inquiry from pollen and seed analysis is that of dietary reconstruction. Increased attention to pollen and seeds will enable archaeologists to expand their current data on meat consumption patterns in the Chesapeake into a more comprehensive interpretation of colonial diet. Although relatively little work has been completed to date on historic sites, some patterns may have already begun to emerge. Something in excess of 10,000 blackberry or raspberry (Rubus) seeds were recovered from the privy (Fig. 48). Steve Mrozowski (personal communication 1996) and Karl Reinhard (personal communication 1996) have reported similar levels of blackberry seeds in 18th and 19th century privies from Rhode Island to North Carolina. Eighteenth-century English cookbooks known to have been sold in Williamsburg, as well as Mary Randolph's 1824 "The Virginia Housewife", list numerous recipes for blackberry wine, raspberry vinegar, and raspberry conserves and jams (Hess 1984). Other seeds from the privy included two species of pear (Pyrus sp.), muscadine grape 61 Figure 48. Seeds from the John Brush privy (from Stephen A. Mrozowski, University of Massachusetts, Boston). (Vitus rotundiflora) and the herb thyme (Labiatea). Other herbs and spices were also represented in pollen spectrum: sage (Artemisia), parsley (Apiaceaea) and mint (Lamiaceae). What makes these spices interesting is their very presence in the privy. Documentary research using probate inventories from the Williamsburg area has indicated that the use of spices was restricted to the elite throughout most of the 18th century (Carr and Walsh 1994). A probate inventory, taken at the time of Brush's death in 1727, valued his entire estate at £ 90, a figure that placed him firmly in the lower middling ranks of society; yet it appears he was using various herbs and spices to season his food. Equally surprising were the large quantities of the mustard family, broccoli and cauliflower are seasonal vegetables, requiring specialized planting and growing conditions. Gardening manuals of the 18th and 19th centuries devoted a great deal of attention to the planting and care of broccoli and cauliflower (Randolph 1826; Langley 1728; Webb-Prentis Papers). Both were vegetables whose use were most likely restricted largely to the wealthy, who could afford the labor it took for the intensive cultivation of these plants.

In addition to their value in dietary reconstruction, pollen and seeds also provide information on landscape reconstruction. Pollen from arboreal species, like the pine (Pinus sp.), oak (Quercus sp), cottonwood (Populus), and sweet gum (Liquidambar sp.) found in the privy, could have been carried by the wind for miles, and thus can be used to assist in reconstructing landscape on a regional, rather than a site-specific level. Low-growing weedy plants, whose pollen movement is somewhat more restricted, can be used to reconstitute the environment at a more local level. Weed species, such as ragweed (low spine Asteraceae), goosefoot (Chenopodiaceae) and pigweed (Amaranth), which grow predominately in disturbed soils, indicate the presence of plowing or garden activity on or near the brush site.

Another form of evidence, although indirect, for the landscape of the Brush property was provided, surprisingly enough, by the parasite remains discussed earlier. Both whipworm and roundworm are transmitted through contact with feces-contaminated soil, which could be expected from eating poorly-washed vegetables grown in a garden that had been fertilized with nightsoil or from poor sanitation practices. Each of the two species of parasitic worms from the privy requires different environmental conditions, with the roundworm flourishing in sunny garden soils, while the whipworm eggs need moist, shady soils for incubation (Reinhard, Mrozowski and Orloski 1986). The much higher numbers of whipworm discovered in the privy suggest that the brush property was shaded by trees in the 1720s.

Two types of seeds recovered from the privy may offer another path of inquiry, although the evidence remains slender at this stage. Five seeds from black nightshade (Solanum nigrum), a weed indigenous to Virginia, were recovered. Although black nightshade and its close relative, deadly nightshade (Atropa belladonna) are poisonous to cattle, black nightshade is listed as having medicinal qualities in "Culpepper's Herbal", dating to the 1670's. The seeds, when mixed with horehound and wine, are listed as a cure for dropsy in another dispensatory. Since the nightshade, like most fruits and vegetables, is an insect-pollinated plant, it most likely entered the privy through being ingested by a human. Seeds from thyme, another plant with medicinal qualities, were also present. Thyme is one of the oldest and most widely used of medicinal herbs, appearing in herbals as early as the third millennium B.C. (Hess 1981). Eighteenth-century sources list it, among other uses, as a pain reliever for decayed teeth. Sage and mint, other ingredients in toothache cures, were also present in the privy. This type of inquiry may shed interesting light into the use of medicinal herbs in the colonial period. Many cookbooks and household guides contained recipes and cures for various ailments, and Gervase Markham, in "The English Housewife" (1615), contained among his attributes of the complete 17th-century woman a knowledge of physical healing and how to administer medicine. This line of inquiry may present an answer for why a middling artisan like Brush had spices and herbs incorporated into his diet. They may not have been used for seasoning his food at all, but for medicinal reasons. Other topics of study suggested by the results of pollen and seed analysis are examining the relationships between diet and health, and how the knowledge of medicinal herbs is transferred through time.

As can be seen from the brief glimpses provided by this analysis, the study of ecofacts from archaeological contexts holds promise for aiding in site interpretation. Although use of pollen from yard or living surface layers can be problematic (see Kelso and Fisher 1989), and even though pollen and seeds do not always survive in the soil (Reinhard 1986), features such as trash pits, garden beds, and wells have great potential for this type of analysis. Questions which before could not be addressed through archaeology, such as those concerning the full range of dietary and landscape reconstruction on local and regional levels, can be studied using this interdisciplinary scientific approach to analysis.

Thomas Barbar and Elizabeth Russell Cary

Very little is known about the owners of Lot 165 and 166 during the years immediately following the death of John Brush. In his will, Brush divided the Palace Green lots equally between his unmarried daughter, Elizabeth Brush, and his son-in-law, Thomas Barbar, husband of Susannah Brush (OW 16:424). By indenture dated February 2, 1727, several months after Brush's death, Elizabeth Brush conveyed her portion of the property to her brother-in-law for the payment of £80 of Virginia currency (DAB 3:440). Barbar, a carpenter, died in May of 1727, leaving his wife Susannah as the executor of his will. She probably continued to live on the property until she sold it, in November of 1728, to Elizabeth Russell, widow, of York County (DAB 3:496-497).

The York County records contain very little information about Elizabeth Russell, but it appears that she was the former Elizabeth Brush, married and widowed within the space of approximately two years. Plans for marriage may have been the reason she initially sold her share of the house to her brother-in-law Barbar. It is probable that Russell continued to reside on the Palace Street lots at least through May 1729, when a slave belonging to neighbor Archibald Blair was accused of having broken into Russell's home and stolen some linens (OW 16:592).

The property was retained directly or indirectly by Elizabeth Russell throughout the 1730s, until its sale to William Dering in 1742. It was probably during Russell's occupation that the kitchen now standing on the lot was constructed in its original form as a 17 x 20 foot frame building on a brick foundation (Savedge 1969). The south end of the kitchen, containing the hearth, was constructed entirely of brick, as a precaution against fire. A bake oven was also built on the eastern side of the hearth. Noìl Hume's 1967 excavations revealed that the original kitchen's walls, fireplace and ceiling were plastered (Frank 1967).

Perhaps it was also during Russell's occupation that the post-supported structure located during the 1989 archaeological excavations on Lot 166 was constructed (Fig. 49). Measuring 15 feet east to west by at least 15 feet north to south, and standing 60 feet north of the kitchen, the building pre-dated the currently reconstructed laundry. Yorktown-type pottery fragments found within the fill of the postholes indicate that the building was constructed after 1720 (Barka 1985), and soil layers sealing the postholes contained ceramics which date the construction before the mid-eighteenth century. It is postulated that this building may have been a stable or a slave house, or both. No evidence of a hearth was discovered, but the entire building was not excavated (Fig. 50).

Between 1729 and 1742 Elizabeth Russell married wealthy landholder Henry Cary II. A clause in Henry Cary's will states that the executor of Cary's will was to pay Elizabeth Cary £220 "in consideration for the like Sum by me heretofore received for the sale of her house in the City of Williamsburg"(Chesterfield County Will Book 1:36-42). Henry Cary II lived in Williamsburg during the 1720s, and evidence shows that he 64 Figure 49. Brush-Everard property as it is believed to have appeared around 1745. Figure 50. Archaeological evidence of possible stable or slave quarter at Brush-Everard. 65 and John Brush were at least acquainted. In 1723, records show the two of them being paid for work they did at the Governor's Palace. Cary, like his father, was a contractor and builder, and was responsible for work on the Powder Magazine, the President's House, and the College Chapel, as well as the Palace. Since his work in Williamsburg continued as late as 1732, it is probable that Cary was residing in town during this period.

Architectural investigations suggest that most of the present woodwork inside the Brush-Everard House was installed during the Cary/Russell tenure (Wenger 1994). New trim included "cornices, chairboards, bases, doors in front rooms, a buffet in the dining room, and virtually all existing woodwork in the passage, including the stair"(Wenger 1994:6). At the same time paint analysis shows that virtually of this trim was painted a medium reddish brown known at the time as "Spanish brown."

By 1742, the same year that William Dering purchased Lots 165 and 166 from Henry Cary and his wife Elizabeth (DAB 5:102-105), Cary is listed as residing in Henrico County. It is possible that Dering was a friend of Cary prior to the land sale, since William Byrd's diary mentions Dering socializing with a Mr. Cary in 1740 and 1741 (Woodfin and Tinling 1942). Another entry for 1741 mentions that Dering visited Byrd at Westover after having come from Mr. Cary's house, and later, visiting Mr. Cary's with William Byrd.

The William Dering Household

William Dering, an artist and dancing teacher, announced the opening of a dancing school at the College in the November 25, 1737 edition of the Virginia Gazette. After purchasing Lots 165 and 166 in 1742, Dering moved into the John Brush house with his wife Sarah and their four-year-old son William. There is no documentary or archaeological evidence that he made any changes to the property when he moved there. Paint analysis of the interior of the house, in fact, suggests that by this time the walls had acquired a thick coat of grime (Wenger 1994).

In 1735 and 1736, according to the Pennsylvania Gazette, Dering instructed students in dancing, reading, and needlework in Philadelphia (Pennsylvania Gazette). He lived in Gloucester County, on the north side of the James, prior to moving to Williamsburg (OW 18:549).

Dering was considered a "Gentleman" by his eighteenth-century peers and was friends with William Byrd of Westover. Byrd's diaries of 1740 and 1741 mention Dering visiting him numerous times on his Charles City County plantation (Woodfin and Tinling 1942). In the company of William Byrd, Dering socialized, played the French horn, and discussed art prints. Dering held no public offices during the time he lived in Williamsburg, but he arranged and held balls and assemblies at the Capitol during Court Days in 1745 and 1746.

Perhaps Williamsburg at mid-century was still not ready for the caprices of entertainment. Like Levingston some twenty years before, Dering began to have financial difficulties almost immediately after purchasing Lots 165 and 166. The York County records indicate that he was being sued for debt by numerous people during the early 1740s (WI: 125, 370). He owed William Prentis, merchant of Williamsburg, a debt of £400, and in 1739, Prentis, George Gilmer and others brought an action of debt against him. In May of 1744, Dering was forced to mortgage his two lots, plus some slaves and personal property to Bernard Moore, Gentleman of King William County, and Peter Hay, Physician, of Williamsburg 66 in order to pay his debt to Prentis (DAB 5:102-105). Dering was to retain use of the property until default on his mortgage.

A schedule of Dering's mortgaged goods, prepared in 1744, show the extravagant lifestyle led by Dering and his family. Among Dering's household goods were five beds, each with assorted bed clothing, numerous chairs, a desk, couch, table, and chest of drawers. Also included were two tea tables, each with their furniture, numerous pictures in gilt frames, china (or porcelain) plates, dishes and bowls, plus kitchen equipment, including pewter dishes, brass candlesticks, and specialized items, such as two coffee mills, one fish kettle and a copper coffee pot. Dering also owned one male and two female adult slaves, plus one slave child. Four horses, plus a chariot, a chaise, and a chair without a carriage listed in the schedule indicate that a stable on the property was virtually a necessity. Perhaps the post structure located in the 1989 excavations on Lot 165 and 166 is the remains of such a stable.

Dering was a man living well beyond his financial means. Although paying off the 1744 debt, he was again forced to mortgage the property in 1745 to Philip Lightfoot of Yorktown, with a default date set for May 11, 1746. At this time, another schedule of Dering's property was prepared. It seems that being in debt did not deter him from spending, since this schedule is more extensive than that prepared just one year previously. Added to the list were a silver watch and spoons, porcelain dishes, table linen and table glass in the specialized forms of jelly glasses and decanters (DAB 5:136-139). Dering never paid off the 1745 debt to Philip Lightfoot, since at Lightfoot's death in September of 1749, his son William Lightfoot assumed the mortgage. Dering still owned a £200 balance of the principal still due, plus £35 for two years unpaid interest, and £35 for an additional advance (Deeds 5:343-345).

By December 1749, advertisements in the South Carolina Gazette show that Dering had moved to Charleston, obviously defaulting on his loan. The Palace Green property was sold at a public outcry on or before February 14, 1751 (Tyler 1899:136). Although Dering continued to advertise in Charleston in November of 1750, there are indirect indications that his wife Sarah may have still occupied the Williamsburg property. For example, the sheriff of York County left a copy of a summons at "the House of the Defendant" on January 21, 1751 (JO I:394). In addition, John Mercer mentions lodging at the Dering house during his visits to Williamsburg in 1749 and 1750.The sale of the property at the outcry recorded in Blair's diary also suggests that Sarah Dering but not her husband was present at the outcry. In any event, William Dering died in early 1751.

Like Levingston before him, Dering encountered financial difficulties trying to make a living in Williamsburg in the arts of entertainment and refined behavior. Williamsburg, although the colony's leading urban center, was still very much a provincial town. The small population it sustained throughout most of the year may not have been able to provide enough patronage to support Dering and his family. Only during Public Times, held four times yearly, did the population of Williamsburg swell to a total of five to six thousand people and provide a market for amusements, such as Dering's balls.

Dering's mortgage schedule shows that he was very much trying to emulate his betters through the acquisition of luxury items like table glass, silver, and horse drawn carriages. Compared with other York County citizens whose estates were valued between £296 and £490, Dering owned more luxury 67 items. Whether the purchase of these items played a part in his financial downfall is not known. Unfortunately, no artifact assemblages which could be definitely linked to the Dering household have been recovered from excavations on the property thus far. There was one excavated soil layer which dated to the second quarter of the 18th-century, which could have been created by any or all of the families living on the property at the time. This layer will be discussed below. Sheet Refuse; Layer 26, a layer of mottled dark grey/brown (10YR4/3) and light tan sandy loam was excavated along the southern edge of Lot 165 and the northern perimeter of Lot 164. It is believed that this layer accumulated as a result of trash disposal and daily activities on these properties during the second quarter of the 18th-century, a period when Lot 164 was owned by George Gilmer and Lot 165 was under the ownership of the several families just discussed. At that time, the boundary line between these two properties was maintained by a fence line. There were no crossmends between the artifacts found on Lots 164 and 165, strongly suggesting that they were produced as the result of trash disposal on different sides of the fence. The archaeological remains recovered from Layer 26 on Lot 164 will be discussed in Chapter 9; the following discussion will pertain only to those archaeological findings from Lot 165.

Layer 26 was located consistently throughout the area excavated south of the Brush-Everard House in 1988 (Fig. 51). A total of 970 artifacts, including at least 29 ceramic and glass vessels (Table 8), were recovered from the 75 square feet that comprised this layer on Lot 165. The nature of this assemblage, with its larger quantities of faunal bone, ceramics, wine bottle glass, and table glass suggest that this layer was built up as a result of household activities. The proximity of the kitchen, located directly to the northeast, helps explain the composition of this assemblage (Fig. 52). The ceramics from this layer include a wide range of decorated delftwares, Chinese porcelains, Westerwald stonewares, English white salt glazed stoneware, and Fulham and Nottingham stonewares. These ceramic types are commonly found on sites from the second quarter of the 18th century. The presence

Figure 51. Location of excavated portions of Layer 26.

A large proportion (over 23%) of the artifact assemblage was window glass. Eighty four percent of this glass and sixty four percent of the nails recovered from Layer 26 were found directly north of the reconstructed garden shed. This may suggest that this building was originally constructed during the second quarter of the 18th-century.

Figure 52. Artifact distribution in Layer 26, Lot 165.

Several other pieces of evidence support the conclusion that this layer was created after the death of John Brush. Layer 26 sealed the privy pit and robbed foundation trench of the privy, thus postdating the John Brush ownership of the property. Additionally, less than one percent of the artifacts (1 gunflint, 3 pieces of coal), recovered from this layer could even remotely be tied with gunsmithing. This compares with the one to five percent ranges associated with Brush period features discussed in the previous chapter.

Archaeological Results

Archaeological excavation on the northern limits of the Gilmer property during 1967 and 1988 revealed extensive features and soil layers dating to the Gilmer occupation of the property. Noìl Hume's archaeological investigations along the northern portion of Lot 164 revealed five trash pits which were associated with Dr. Gilmer's house and shop. Portions of some of these trash pits, left unexcavated in 1967, were re-examined and dug in 1988. A sixth trash pit was also discovered and excavated at this time. The large quantities of delft drug jars, salve pots, and pharmaceutical bottles contained within the pits constitute the finest assemblage of mid-eighteenth century apothecary-related items ever recovered in a Williamsburg excavation. These features will be described in detail below.

Gilmer's Apothecary Shop

Apothecary shops, primarily concerned with the preparation and sale of drugs, were beginning to be established in larger towns throughout the American colonies by the beginning of the eighteenth century (Gill 1972:30). Williamsburg, as the capital of the colony, was apparently able to support several apothecaries at any given time. Gilmer, a surgeon as well as an apothecary, was among the first to operate a shop in town. During Gilmer's ownership of the shop on Palace Green, other apothecaries were being run by Thomas Wharton (ca. 1735-1746), Dr. Peter Hay (1744-1766), and Dr. George Pitt (1744-1768). Despite his competition, however, Gilmer was apparently very successful.

Although never located archaeologically, Gilmer's shop was probably very similar to other documented eighteenth-century examples. Most shops contained several rooms: one for the display and sales of merchandise, one for consultation, and another for a laboratory. Large quantities of drugs and medicines were typically imported several times a year from England (Gill 1972). In addition to selling medicine, Gilmer also carried a large stock of groceries and spices, which he advertised regularly in the Virginia Gazette. His stock included anchovies, pistachio 77 nuts, mushrooms, vermicelli, coffee, cinnamon, cloves, tooth powder, prunes, currants, and sugar, along with other spices, fruits, and teas.

Many of the ceramic and glass vessels recovered from the trash pits and soil layers originated in Gilmer's apothecary shop. The most prevalent ceramic forms, delft drug jars and ointment pots (Fig. 55), were cylindrical vessels for powders, pills, salves and confections (Garner and Archer 1972:14). Undecorated examples, which were inexpensive to produce, outweighed the decorated jars and pots three to one. In fact, some of the undecorated examples had been discarded unbroken, suggesting that they were perceived as disposable items in the eighteenth century. This did not appear to the be case with the decorated drug jars, represented by examples painted in blue with combinations of linear, cross and chain patterns. These were invariably found broken, suggesting that the decorated examples, more expensive to produce, were recycled in Gilmer's shop until they were no longer of any use. Analysis of the painted drug jars revealed that they were old at the time of discard, perhaps even having been brought to the colonies by Gilmer in 1731 (Beck and Pittman 1992).

Fragments of two undecorated delftware pill tiles and two cylindrical delftware storage jars were among the pharmaceutical materials recovered. Pill tiles, sometimes marked with a graduated scale, were used to roll and cut medicines into individual pills (Griffenhagen 1957:14). Not present in the archaeological assemblage were any of the large elaborately decorated spherical drug jars more commonly depicted in contemporary prints of apothecaries. These ornate vessels were used primarily for decoration (Gill 1972:67) and would not have endured the daily handling that contributed to the higher breakage rate of the smaller, less decorative drug jars and ointment pots.

Figure 55. Examples of drug jars and salve pots from Gilmer's trash pits.

In the June 20, 1745 edition of the Virginia Gazette, Gilmer advertised Hungary Water, Spaw Water and Pyrmont Water. Fragments of sixteen Pyrmont water bottles were recovered from Trash Pit A. These bottles, which had contained mineral water from Germany, were marked with a distinctive seal, molded with a crowned shield of arms.

Trash Pit E contained a human lower mandible identified as that of an elderly African-American female (Frank 1967:36). It is suspected that this mandible was part of a skeleton used by Gilmer for study, but how and where Gilmer obtained it is unknown.

The Gilmer Household

After purchasing the Palace Green lots, Gilmer almost certainly moved into the house constructed between 1716 and 1718 by William Levingston. Although excavations of the house in which he resided showed no evidence that Gilmer enlarged the house in any way, documents show that he renovated the interior in the early 1750s. The interior appearance of Gilmer's house is suggested by the letter he wrote to merchant Walter King in August 1752, in which he states: 78

Mrs. Gilmer is perfectly satisfied with your conduct about her China.... I have just finished a closet for her to put it in as agreed on before you left us. I am wainscoting my dining room, which with a handsome marble chimney piece &c with glass over it, will make it a tolerable room for an Apothecary(Brock Notebook).

Gilmer refers in detail to the interior renovation of his house. One of the recorded items was the installation of a marble chimney piece. Perhaps this stone replaced a fireplace surround of delft fireplace tiles, of which two decorated examples were found in trash pits dating to his occupation.

The results of Gilmer's renovation may be evident today in the parlor of the Tucker-Coleman house, whose core was formed by moving the Levingston/Gilmer house to Lot 169 in the late eighteenth century. The parlor contains eighteenth-century wainscoting, carved chair rails, cornices, and baseboards. The original corner cupboard in what is today the parlor of the Tucker Coleman House may represent Mrs. Gilmer's china cupboard. The Brock Notebooks also reveal that the interior of Gilmer's house was furnished with mirrors, status ceramics, and a desk. Additionally, he purchased two paintings from the estate of John Collett in 1751 (OW 20:215-217) with which to decorate his home.

Figure 56. Chinese porcelain plate from one of Gilmer's trash pits.

In the previously-quoted letter to merchant King, Gilmer mentioned the set of china that was being sent to Mrs. Gilmer, and the closet constructed to hold the porcelain (Brock Notebook). Thirty-three porcelain vessels, including two matching blue and white Chinese porcelain plates, were recovered from the trash pits (Fig. 56). This porcelain, which contains an unusual central motif and detailed painting, is a high quality ware, whose shape and painting style place its production in the first quarter of the eighteenth century. Due to their earlier manufacture date, it is doubtful that these were the plates ordered by Gilmer in 1752. It is more likely that fragments of overglaze porcelain decorated in red and black also recovered from the trash pits represented this order. The Gilmer family also owned Chinese porcelain teawares in the forms of cups, saucers, tea bowls, and slop bowls. Tableware forms included ten plates, a platter, and three punch bowls, with Chinese porcelain making up twenty-three percent of the Gilmer household ceramics. Seventy-six percent of the porcelain recovered was contained within the latest of the five trash pits, perhaps suggesting that Gilmer's expenditures on ceramics and other household goods rose towards the end of his life.

In addition to the porcelain, numerous ceramics of English manufacture were 79 found in the trash pits. Two molded white salt glazed stoneware plates, first manufactured after 1740 (Noìl Hume 1969:115), were recovered. Teaware forms were also found in white salt glazed stonewares and decorated English delftwares. While delft tablewares were distributed evenly throughout all of the five pits, the only delft teaware vessel found was contained within an early trash pit. Although delft was used in the eighteenth century as a low cost substitute for Chinese porcelain, it was not particularly effective as a teaware. The fragility of the glaze could not withstand repeated contact with hot liquids without cracking and breaking (Garner and Archer 1972:22). The scarcity of delft teawares and its position within the earliest trash pit indicates that Gilmer quickly replaced his delft teawares with more durable and more costly English white salt glaze stoneware and Chinese porcelain.

Analysis of the Gilmer ceramics indicates that the family was eating from blue and white painted delft tablewares during their first years of lot ownership, replacing these with molded white salt glaze plates and other tablewares sometime after 1740.

Summary

The documentary and archaeological evidence portray Gilmer as a man of middle age, successful both in business and public service. Part of this success is evident in his entrepreneurial activities, which consisted largely of purchasing rental property, such as his joint venture in the Raleigh Tavern (DAB 5:495-497). Gilmer accumulated enough capital to send his son to Europe to study medicine (Gill 1972:55). His letter to King shows that he was not only concerned with outward appearances of wealth, but that he felt he was successful. John Blair's diary records that Gilmer entertained the Governor at his home upon several occasions (Tyler 1899:148). He also maintained a close friendship with other gentry, including John Blair, entertaining him, corresponding with him, and eventually marrying his sister, Harrison Blair, in 1745.

This concern for outward appearances surely played a part in Gilmer's decision to purchase expensive Chinese porcelain tablewares, and to build a special cupboard for their display in one of the more public rooms of his house. Even at mid-century, Chinese porcelain was still a luxury, with only 50% of all households in the highest wealth category owning specialized drinking and dining equipment such as porcelain (Carson and Walsh 1981). In addition, evidence from the trash pits indicates that some of Gilmer's porcelain was part of a matched set of excellent quality. Sets of ceramics did not become common until the nineteenth century, another testament to Gilmer's concern with ownership of status items. Some 76% of Gilmer's porcelains (19 vessels), in both table and teaware forms, were blue underglaze handpainted in floral and Chinoiserie designs. Six vessels decorated in overglaze handpainted designs, in forms equally divided between tea and tableware, were found in the uppermost layer of the latest trash pit. This seems to indicate that this porcelain, which represented the most costly ceramic items found in the trash pits, was among the latest of Gilmer's ceramic acquisitions. These overglaze porcelain pieces, which show designs common to the second half of the eighteenth century (Noìl Hume 1969:259), may have been part of the china ordered in 1752. Additionally, over three-quarters of the porcelain found in the Gilmer assemblage was from the latest trash pit, indicating that the bulk of his porcelain acquisition occurred during the latter portion of his occupancy. Only eight porcelain vessels were recovered from the four earlier 80 trash pits, and these forms were decorated using the less expensive technique of blue underglaze handpainting.

Gilmer continued to live on the property until his death in January of 1757, leaving the property to his son, Peachy Gilmer (WI 20:423).

Renovations During Everard's Tenure

Documents and archaeological evidence suggest that by the early 1770s Thomas Everard performed extensive renovations to structures on the property. Everard's initial work on the house in the first decade of his tenure seems to have consisted of applying its first coat of white paint (Wenger 1994). But between 1769 and 1773 Everard began purchasing large quantities of nails, glass, and paint. He appears to have installed the first-floor chimneypieces and wainscoting, the wallpaper, and the mantelpiece and paneling over the door in the south wing (Wenger 1994). Scaffolding holes were found on the east side of the north wing during the 1994 waterproofing excavation (Pickett 1994), indicating the some exterior repairs on this portion of the house were undertaken at this time as well.

It was once believed by Colonial Williamsburg researchers that Everard added the two eastern wings to the house. Recent dendrochronology, however, shows that the north wing at the very least was constructed soon after 1720 (Heikkenen 1984) by original property owner John Brush. Unfortunately, the nature of the 1947 archaeological excavations and the subsequent rebuilding of the south wing make dating its construction impossible. Although it is possible the construction of the south wing was contemporary with that of the north wing, this addition could just as easily have been built by one of the property's later residents: Elizabeth Russell, Henry Cary, or William Dering. It is interesting to note that Henry Cary's circa 1732 design of Ampthill in Chesterfield County was constructed with the same floor plan as that of the Brush-Everard House (Kocher and Dearstyne 1950), suggesting that the Williamsburg house had achieved this floor plan before or during Cary's tenure there in the 1730s.

The methods of the 1947 archaeological trenching also make it impossible to determine the construction dates for the reconstructed laundry, office and original smokehouse. Fortunately, better dating information exists for the kitchen. Ivor Noìl Hume's 1967 excavation of this building revealed that it was constructed in three phases. Originally built after 1730 as a frame structure with a packed clay floor, the kitchen (Fig. 58) was altered to become a brick structure in the mid-eighteenth century (Frank 1967). A layer of destruction debris associated with this renovation showed that the first period kitchen contained plastered walls and ceiling (Frank 1967:23). A fragment of delftware plate matching one found at Everard's earlier Nicholson Street home was found within the renovation debris, indicating that Thomas Everard was responsible for converting the kitchen to a brick building. The northern addition to the kitchen was probably constructed around the mid-eighteenth century, possibly as late as 1790 (Frank 1967:18). When this addition was built, three dormer windows were added along the west roof, increasing the light and livability of the space above the first floor (Savedge 1969). This renovation may be related to Everard's need to house the increasing numbers of slaves residing on the property by the 1770s. The laundry building, directly north of the 84 Figure 58. Periods of kitchen development at Brush-Everard. kitchen, and its mirror image, may have been constructed by Everard when he converted the kitchen to brick. In addition to its primary function as a laundry, this building could have also served as slave housing.

Combining archaeological and documentary evidence suggests that Everard's renovation of the property did not begin until the late 1760s. Between 1769 and 1773, Everard ordered 40,000 nails, 100 feet of window glass and 100 pounds of white lead paint from John Norton, his merchant in England (Norton Papers). Window glass, ceramic paving tile, mortar and plaster fragments, delft tiles, leaded window cames, and iron door brackets (as well as countless wrought iron nails) found in both trash deposits from the Everard period also attest to this period of renovation and/or new construction. While archaeological evidence suggests that dependencies around the house were built or renovated, some of these changes apparently took place within the dwelling. It is likely that Everard was responsible for adding a new staircase to the front hall of the house, whose elaborate floral carving is reminiscent of that at Carter's Grove and Tuckahoe plantations. The curved gooseneck rail end on the stair and the fine panelling in the stair hall are also atypical of houses of the size and construction of Brush-Everard.

When architectural restoration took place in 1949, it was found that the downstairs rooms in the original 1718/19 house were fairly plain and undecorated. Later additions, probably by Henry Cary, of carved chair rails, fireplace surrounds, baseboards, panelling, and six panelled doors converted the downstairs rooms of the house into more elaborate spaces for entertaining guests or clients (Wenger, personal communication, 1996). Other documentary evidence indicates the elaborate nature of Everard's downstairs, or public, rooms. The wallpaper found in the library during restoration possibly dates from the mid- to late eighteenth century (Pritchard 1985). Orders to merchant John Norton reveal that Everard's home contained upholstered furniture (Norton Papers) and he is also known to have owned a chess set and a backgammon 85 table, purchased from the estate of Governor Fauquier in 1772 (WI 22:83-89). These elaborate furnishings and architectural detailing would befit a man of Everard's status and position in the Williamsburg community.

The differences between the public and the private spaces within the household are very evident in the different architectural treatments of the rooms. The second story rooms, which were used as sleeping areas during Everard's occupation, contained no cornices, chair rails or mantels, and only simple, undecorated baseboards. It was probably during Everard's occupation, at the same time the stairway was added, that dormer windows were cut through the roof, allowing the second floor of the house to become a lighted, useable space.

Everard Ravine Deposits

The 1987/88 archaeological excavations on Colonial Lot 166 revealed that Everard continued the pattern of garbage disposal established earlier on the lot by Brush. Everard's household deposited a great deal of garbage, mainly in the form of kitchen and food-related debris, into the natural ravine at the north end of the property. Two large deposits of ashy soil comprised the predominant assemblages dating to Everard's tenure. Several additional soil layers, also in the ravine area, were formed during the third quarter of the eighteenth century (Fig. 59). Because of later disturbances, no soil layers dating to the Everard period were evident around the house or in the pasture area of Lot 166, although at least one deposit of renovation debris was present around the colonial kitchen (Frank 1967). Following will be descriptions of each soil layer and feature associated with the Brush-Everard property.

Figure 59. Profile through ravine at north end of Lot 166.

Main Ash Ravine Layer
(29F-69, -73, -82, -112)

The most artifact-laden layer on the site (Table 12) was a deposit of dark reddish brown (5YR2.5/2) sandy loam in the ravine at the northern end of the property (Fig. 60). This deposit of ashy soil, ranging from 0.2-0.55 feet 86

Table 12.
Ceramic and Glass Vessels from Main Ash Layer

Type	Form	No. of Vessels	% of Total
Delftware	Plate	19	7.0
	Dish	6	2.2
	Soup plate	2	0.7
	Bowl	4	1.5
	Punch bowl	12	4.4
	Basin	2	0.7
	Drug jar/ointment pot	24	8.9
	Chamberpot	6	2.2
	Tile	10	3.7
	Hollow	1	0.4
Westerwald stoneware	Chamberpot	9	3.3
	Jug	1	0.7
	Mug	3	1.1
White salt glaze stoneware	Bowl	3	1.1
	Patty pan	1	0.4
	Teapot	2	0.7
	Saucer	1	0.4
	Can/mug	3	1.1
	Tankard	1	0.4
	Chamberpot	1	0.4
	Soup plate	1	04.
	Cup	1	0.4
	Cream jug	2	0.7
Dipped white salt glaze stoneware	Tankard	3	1.1
	Bowl	1	0.4
Chinese porcelain	Plate	7	2.6
	Saucer	24	8.9
	Cup/can/tea bowl	22	8.1
	Bowl	4	1.5
	Dish	1	0.4
English porcelain	Bowl	1	0.4
Creamware	Bowl	1	0.4
	Plate	1	0.4
	Teapot	1	0.4
Agateware	Teapot	1	0.4
Astbury	Teapot	1	0.4
	Mug	1	0.4
	Bowl	1	0.4
Wedgwood	Bowl	1	0.4
North Midlands slipware	Mug	1	0.4
	Flatware/plate	1	0.4
	Storage jar	14	5.2
	Tankard/can	2	0.7
	Pipkin	1	0.4
	Jug	1	0.4
Nottingham stoneware	Mug/tankard	9	3.3
	Bowl	3	1.1
	Patty pan	1	0.4
	Pitcher	1	0.4
	Hollow	1	0.4
American stoneware	Storage jar	1	0.4
Colonoware	Pipkin	1	0.4
87
	Hollow	1	0.4
Yorktown-type earthenware	Milkpan	8	3.0
	Bowl	1	0.4
	Bird bottle	2	0.7
	Pipkin	2	0.7
	Chamberpot	1	0.4
Buckley earthenware	Storage jar	3	1.1
	Butterpot	1	0.4
White sandy earthenware	Storage jar	1	0.4
	Milkpan	1	0.4
Unspecified coarseware	Bowl	2	0.7
North Devon earthenware	Milkpan	1	0.4
Iberian earthenware	Storage jar	12	4.4
Colorless leaded glass	Stemmed glass	8	3.0
	Decanter	2	0.7
	Tray / salver	1	0.4
	Jelly glass	1	0.4
	Tumbler	1	0.4
Colorless leaded glass	Handle	1	0.4
Total		271	100.00

Figure 60. Wine bottle glass and coarse earthenware in situ in Everard main ash layer. 88 in thickness, sealed the brick and mortar renovation layer of the John Brush period. Based on artifact dating, the deposition of this layer occurred between the beginning of Everard's tenure in the 1750s and about 1770. While not completely excavated, this deposit was tested extensively. Some 480 square feet of soil from this deposit were excavated, with a total of 18,660 artifacts and 271 ceramic and table glass vessels recovered. Since it was noted, as excavation proceeded, that preservation of animal bone and other food remains appeared to be excellent, the standard sampling procedure for faunal and other food remains was altered. All soil from this layer was passed through one-quarter-inch mesh screen into an one-eighth-inch mesh screen. Any soil which did not pass readily through the smaller screen was bagged for later flotation and waterscreening through window mesh screen. This facilitated the recovery of many small, otherwise easily missed fragments such as eggshell, seeds, crab claws and bones from birds, small mammals and fish. With this sampling strategy, a more complete picture of the Everard household diet can emerge. In addition to the microfaunal remains, many smaller artifacts which usually pass unnoticed through one-quarter-inch screen, such as straight pins and glass beads, were recovered.

Everard and His Contemporaries

Comparing Everard's ceramic assemblage with those of some of his Williamsburg contemporaries suggests that his table and teaware was fairly typical for the time period and cut across social and economic levels. Everard was compared with fellow gentryman Peyton Randolph (assemblage dating 1755-1765), John Draper, a middling blacksmith (1769-1780 deposit), and the up-and-coming tailor and merchant Robert Nicolson (deposit dated 1751-1797) (Edwards et al. 1988:36; Brown et al. 1990:238-239; Samford 1986:45). Despite the introduction and subsequent popularity of refined earthenwares and stonewares during the second and third quarters of the eighteenth century, delft tablewares continued to play an important role in all four assemblages, particularly as decorated plates and punchbowls. Molded white salt glaze stoneware plates, beginning manufacture around 1740, were also prevalent in the other three assemblages, as were porcelain table 97 and teawares, both in underglaze and overglaze decoration. Even middling artisan John Draper owned table and teawares in porcelain, one of the most expensive ceramics available during the eighteenth century. Overall, there seemed to be little difference between the four households in terms of types of ceramics owned.

Recent work has focused on the beginnings of courtly traditions of cooking and serving food (Yentsch 1990) and the burgeoning interest in individualization as the eighteenth century progressed (Deetz 1977). This new tradition took the form of specialized food preparation and serving items. As for ceramic tablewares, plates for serving individuals became much more common as the use of communal bowls, cups and dishes began to disappear. With this emphasis on each diner receiving his own plate and cup, the use of other individual utensils also increased. While forks were still uncommon in Virginia stores in 1745, by 1770 both forks and knives were readily available in a variety of prices and materials (Martin 1987).

Recently, archaeologists have sought economic patterning in the types of ceramic wares owned by households of different wealth categories (Adams and Boling 1989; Miller 1974; Otto 1984). While it does appear that certain wares were restricted to wealthier households in the seventeenth and early eighteenth centuries, the Williamsburg results outlined above suggest, perhaps, that ownership of certain types of ceramics was less relevant than other factors. Wealth may have been expressed instead by either the sheer numbers of vessels owned or in the types of vessel forms purchased.

The ownership of specialized vessel forms, which went along with the preparation and presentation of fashionable food, may be a better indicator of wealth and social standing. The proliferation of cooking and serving equipment which became available during the eighteenth century reflects the development of a more elaborate cuisine, one that was focused on numerous food courses served on individual place settings at cloth-covered tables. Ceramic and glass vessels found in Everard's garbage, as well as documentary references indicate that the mayor was participating in high-style dining and entertaining, using the mealtime display as an indicator of his position within the social hierarchy of Williamsburg (Yentsch 1990). His order of a venison pasty pan from his London merchant in 1773 (Norton Papers) and fragments of an iron reflecting oven found around the kitchen indicate that he was preparing elaborate game pies as well as roasted meats. Everard's ability to keep his tea in a cauliflower-shaped caddy instead of a plain wooden box, and the glass salver adorned with jelly or wine glasses, are more symbols of wealth and social standing than they are items necessary for the daily operation of a household. The appearance of these specialized vessels on Everard's table would have communicated to his guests that their host was conversant in the language of fine dining, at a time when, as Lorena Walsh has stated, "eating was ... as much a social event as a means of satisfying hunger." Everard's assemblages contained a profusion of evidence that he was able to entertain in high style, certainly a necessity for someone of his political and social station. Guests could be served from plates on a table adorned with glass and ceramic vessels in expensive, specialized forms. Doubtless this table would have been covered with a tablecloth and each guest would have received a cloth napkin to protect their clothing. Knives, forks and spoons found in the ravine garbage show that Everard's guests would not have been required to eat with their fingers.

Archaeological evidence reveals not only the utensils and tablewares from which 98 Everard and guests were eating, but also the types of food served. Everard's meat diet was quite diverse, with a broad range of both domestic and wild animals. Typical of most archaeological assemblages from eighteenth-century Virginia were large domesticated mammals, such as cow, pig and sheep, which made up the largest percentage of the animal bone assemblage. The diet was supplemented, however, by a variety of fresh and salt water fish (sturgeon, Atlantic cod, striped bass, black drum, catfish, sheepshead and white perch), wild game (deer, quail, Canada goose, various species of duck) and domestic fowl (chicken, goose, and turkey). Water-screening and flotation revealed the skeletal remains of smaller animals, such as blue crab, as well as eggshell.

Everard, with his numerous resources, could have obtained food for his table in a number of ways. By this period, there was an established market system in Williams-burg, where meat and produce could be secured at least twice a week, and probably daily. Some food was evidently supplied directly from the Palace Green lots: Everard advertised cattle missing from his Williamsburg property in the late 1760s. Since Everard was keeping cattle at his home, he almost certainly was also raising domestic fowl-chickens and geese. The discovery of an iron hoe from soil layers around the kitchen suggests a kitchen garden supplying fruits, vegetables and herbs.

Although evidently a widower throughout most of his stay on Palace Green, Everard did have two daughters, Frances and Martha, living with him at various times. While documentary references to his daughters were rare, there is little doubt that they would have exerted influence in the household. The tasks of instructing slaves, entertaining guests and planning meals were just a few of the tasks which fell to women in colonial households. Everard's daughters, in the probable absence of a mother, would doubtlessly have assumed many of these roles while they lived with their father. Their influence in choosing household furnishings such as ceramics and furniture, or in deciding what foods the family and guests would eat, is unknown.

Frances Everard apparently suffered from chronic ill health. After being widowed, she lived with her father during the last year of her life, finally dying in 1773 at the age of twenty six (Gibbs 1984). It is believed that the number of drug jars, ointment pots, and glass medicine bottles found in Everard's garbage assemblages may be related to Frances Everard's illnesses.

Summary

Although Everard continued as the clerk of York County until his death in 1781, he began to give up public service positions as early as 1776. In that year he resigned from the Court of Admiralty, and in 1780 he gave up his position as auditor (VCP/VSL). Thomas Everard died between January 31 and February 19, 1781. No will or inventory of his estate has been located.

Documentary and archaeological evidence indicates that Thomas Everard rose in power and wealth during the years he lived on Palace Green. His move to Lots 165 and 166, adjacent to the Governor's Palace, were indicative of his upward mobility, providing his family with a more prestigious location in town, as well as a larger house. As his political power increased from official positions at the local level to colony-wide offices in the years preceding the Revolution, Everard's economic wealth also rose. Some of this wealth was expended on property renovations and personal possessions which served as signals to others of his position in Williamsburg and Virginia society. The carved staircase of the front entrance hall, rivalling that of the Carter's Grove entrance hall, was a bold statement about Everard and his position as a Gentleman in Virginia society. Another social statement would have been made when visitors to Everard's house were greeted by liveried slaves.

1967 Findings and Preliminary Documentary Research

During the 1967 archaeological project at Brush-Everard, excavation south of the reconstructed garden shed revealed traces of a structure. Sealed beneath a thin layer of topsoil were a brick hearth base bonded with lime mortar, as well as two associated brick piers. A fragment of pearlware with a beginning manufacture date of around 1800 was discovered in the soil layer underlying the brick hearth, indicating that this building had been constructed some time after that date. In the archaeological report on these excavations, it was concluded that these structural remains represented a pier-supported structure built during the nineteenth century (Frank 1967). The full extent of this building was not traced at this time.

During preparation of a documentary and archaeological summary of the Brush-Everard property prior to beginning the 1987 excavation, an early twentieth-century map of the adjacent Tucker-Coleman property was examined (Williamsburg Restoration 1928). Depicted at the northern end of the Tucker-Coleman property, abutting the Brush-Everard lot, was a schematic representation of a structure labeled "Building stood here in 1800 - Home of Colored Mammy and her Husband." Overlaying a scaled version of this map with the structural remains located during the 1967 archaeological project strongly suggested that the archaeological and map data represented the same building. Since it had been established during the 1967 project that the actual boundary between the Brush-Everard and Tucker-Coleman properties was approximately 20 feet north of the presently reconstructed fenceline, this land in fact belonged to the Tucker family when the pier structure was standing. Additionally, the 102 memoirs of John Charles, recalling Williamsburg on the eve of the Civil War, stated that there "was the old frame house which stood on the western line of the `Tucker' lot with side to the street. This was used as sevants' quarters"(Charles 1933). Because the documentary evidence seemed to indicate that this building could be a slave house, the Department of Archaeological Research leapt at the opportunity to excavate an urban slave house, albeit one that dated to a slightly later time period than that which Colonial Williamsburg normally studies.

The documentary record, in this case, is quite good. The 1928 map was drawn by George P. Coleman, grandson of Nathaniel Tucker (Tucker 1942:19), and shows every indication of being accurate. Additionally, the Tucker-Coleman papers, housed at the College of William and Mary, contain a wealth of information about the Tucker family slaves, which was seen as a potential source for information about this structure and its inhabitants.

The Tucker Family

St. George Tucker, professor of law at the College of William and Mary and federal judge in the Virginia district court, built his home along Nicholson Street in the late eighteenth century. One of his sons, Nathaniel Beverley Tucker (1784-1851), also studied law, practicing in Charlotte Court House and Roanoke, Virginia during the early years of the nineteenth century (Sturzenberger 1979:28). In 1815, Nathaniel moved with his first wife and family to Missouri, where he served as a circuit judge until the 1830s. In 1833, he accepted a law professorship at William and Mary, bringing with him his third wife, Lucy Anne Smith, and the beginnings of their new family. They moved into his father's former home, where direct descendants of the family still continue to live.

Polly Valentine

Documentary research, primarily in the Tucker family papers, showed that the most likely resident of the structure at the north end of the Tucker property was a woman named Polly Valentine. Her owner, Nathaniel Beverley Tucker, had received Polly from his half-brother, John Randolph of Roanoke, as part of a larger group of slaves. Polly Valentine served for many years as a nanny to the Tucker children, and family records indicate that she was a favored family slave.

Today's researchers are fortunate enough to have been left a relative wealth of information about Polly Valentine and her family. Most of this information comes from reminiscences of Cynthia Beverley Tucker (1832-1908), daughter of Nathaniel and Lucy Tucker, recalling life in Williamsburg in the early 1840s. Tucker described "Mammy Polly" as "a tall well-made woman" whose head was "not covered with the conventional turban, but combed and bound with yards and yards of narrow black braid" (Coleman 1934:55). Polly was married first to Robert Christian, a neighbor's slave. At night, Christian would stay with his wife in her "house," decried as "a neat comfortable room" in the laundry, where the other family slaves also resided (Coleman 1934:54). According to Cynthia Tucker, her mammy's first husband had a drinking problem and was often cruel. Upon Polly's urging, Nathaniel Beverley Tucker sold Christian in order to release her from this marriage. In time, she married Jim Valentine, in a service held in the vestry room of Bruton Parish Church and attended by Cynthia Tucker and her siblings. After their marriage, Valentine "came to live at her [Polly's] home and, as children grew up around them a house was fixed up for them apart from the other servants" (Coleman 1934:56). 103 It is believed that the house excavated in 1988 by Colonial Williamsburg archaeologists represents the home Nathaniel Beverley Tucker had constructed for the Valentine family. At present, this is the only known excavated structure built expressly to house slaves in Williamsburg during the eighteenth or nineteenth centuries.

Archaeological Results

One of the goals of the excavation at Brush-Everard during the summer of 1988 was to uncover any further remains of the house believed to have been occupied by Polly Valentine and her family. Excavation encompassed the area which Noìl Hume's crew had explored around the slave house in 1967, as well as extending to the west, north, and south. Consistent with the results of the earlier excavation, the structural remains for the house were encountered underneath a thin layer of topsoil. Although the building's hearth base had been removed during the 1967 investigation to allow the excavation of several trash pits underlying the brickwork, the two previously-discovered brick piers remained intact. Three additional brick and stone piers were also uncovered, although modern disturbances, such as utility trenches, had cut through areas believed to have once held additional house piers. Excavation revealed the remains of a 15 x 25 foot wood frame structure resting on brick and stone piers, and containing a substantial brick fireplace at its eastern end (Fig. 66).

Figure 66. Archaeological remains of the Polly Valentine house.

The archaeological remains uncovered at the Valentine site are consistent with prescribed slave housing of the mid-nineteenth century (McKee 1989; Singleton 1988; see Table 16). Larry McKee's research has shown that antebellum writers championing orderly slave life in southern agricultural journals recommended that slave housing be constructed to fit specific requirements. It was thought that by living in permanent, comfortable, single family homes, slaves 104

Archaeologically, it appears that the Valentine house was destroyed sometime around the Civil War, or shortly thereafter. Documentary references to the Union occupation of Williamsburg indicate that many wooden outbuildings in town were demolished by Federal troops and used for firewood (Charles 1933). Since Union soldiers inhabited the Tucker house during the occupation of the town, perhaps this fate befell the Valentine house. Cynthia Tucker's memoirs indicate that most of the Tucker family slaves relocated to Boston and other areas after the close of the Civil War. Polly Valentine, who apparently never left the state of Virginia, later returned to Williamsburg to nurse Lucy Ann Tucker through the illness that eventually resulted in her death.

Discussion

Excavation of soil layers under and around the house provided an assemblage of artifacts which are believed to have been associated with the Valentine family (Table 17 [from Edwards 1990:Table 1]). A recent study of the Polly Valentine site examined material possessions as an indicator of master-slave relations on the Tucker property in the mid-nineteenth century (Edwards 1990). Edwards hypothesized that the slaves were given what the Tucker family considered good treatment (prescribed housing, more than adequate material possessions and food) in exchange for not granting the slaves freedom. The "fit" of the Valentine house was compared with the prescribed standards, as well as an analysis of the ceramic vessels excavated from soil layers around the house.

Several generations of the Tucker family owned African-American slaves who lived and worked on the lots overlooking Market Square. While dependent on these slaves for the smooth operation of his family's daily lives, as well as the status they helped confer on the family, St. George Tucker had mixed feelings about the institution of slavery. In his Dissertation on Slavery (1796), Tucker advocated granting slaves freedom. Unlike his father, Nathaniel Beverley Tucker, author of numerous proslavery writings, was a full supporter of slavery. Typical of many antebellum slave holders, he saw slaves as childlike and dependent upon the care of their white owners. This paternalistic attitude was very evident in his treatment of the slaves (always referred to as servants), which he owned both in Williamsburg and on his other properties. Family letters often included fond greetings to the slaves and news about their family members on other plantations. One telling example of Tucker's paternalistic behavior was his willingness to sell Polly Valentine's first husband in order to free her from an unhappy marriage.

Interpersonal relationships between owners and slaves varied by situation, but were particularly intense for slaves working directly in the homes of, or under the direct supervision of, their owners. Contact was frequent, and in the best interest of the owner, at least, was hopefully placid. In return for duties and obligations performed, masters often reciprocated with incentives or rewards. In some cases, slaves may have even come to expect such gratuities or "kindnesses." For example, Usa Payton, one of Nathaniel Tucker's slaves living in Missouri, wrote in 1851 to her master, updating Tucker on her family and reminding him that he had "all ways [sic] treated me well & I shall forever like you for it...."(Coleman 1934:46). She continued by asking him to send her a present of money, since hard times had fallen on the St. Louis household, requiring her to take in outside washing. Whether Payton's request was granted is unknown, but the fact that she felt she could make this request is telling. Several other Missouri slaves petitioned Nathaniel Tucker not to separate them from their families by sending them to Texas (Coleman 1934:39). A list, prepared several years later, of slaves on the Texas plantation does not list the names of either woman, so it appears likely that the women's request was granted.

Historical documentation suggests that slaves in urban places were distributed in various locations on in-town properties, both inside the owner's homes and in various outbuildings. In 1835, many of the outbuildings on the Tucker property, including slave housing, caught fire and burned to the ground (Tucker-Coleman Uncat. Folder 337). Descriptions of this fire mention a strong wind blowing from the southwest and that the Brush-Everard house was at great risk of being destroyed, suggesting that the burned outbuildings had been arranged on the northern portion of the Tucker lot. Although the buildings had been insured, the Tucker slaves were left temporarily homeless, and possibly without possessions. Despite Nathaniel Tucker's assessment that "nothing was lost," this would have surely been a gross overstatement to the slaves who probably lost what material possessions they did own. Perhaps it was at this time the slaves began to be housed in the laundry, where Valentine resided with her first husband. Martha Vandegrift, interviewed in 1932, recalled Williamsburg slaves and their housing around the beginning of the Civil War and, more specifically, the Tucker family slaves: "They lived out in yards. All the kitchens were outside.... Mrs. Tucker's servants had nice comfortable homes. They were well lodged. They lived on the place. 107 They were excellent servants."(Vandegrift 1932). Vandegrift's statements seem to imply that, while most slaves lived in kitchens, the Tucker family slaves had houses separate from other outbuildings on the property. This may have been one result of Tucker's paternalistic attitude towards the slaves who attended to the family's daily needs.

Certainly Polly Valentine was one of the more favored of the family slaves, entrusted with the responsibility to caring for the Tucker children. Valentine was certainly spoken of with love and affection by Cynthia Tucker, and actions taken on her behalf by Nathaniel and Lucy Ann Tucker indicate their strong, positive feelings as well. Her housing would have been considered more than adequate for the time period and its location at the extreme northern end of the property guaranteed some degree of privacy from the Tucker family.

The Documentary Record

John Brush was one of Williamsburg's earliest gunsmiths, arriving some time before July of 1717. He had received his training in London, where he had been a member of the gunsmith's guild. Brush was probably brought over by Lieutenant Governor Alexander Spotswood to fill the colony's desperate need for skilled metalworkers.¹

Between 1717 and 1723, he apparently made and serviced guns for both the Governor and for private citizens. Spotswood established the Magazine at Williamsburg in 1715 and relied on Brush to maintain and repair guns belonging to the colonial government. These arms supplemented the Virginia militia. He also apparently made pistols and pieces for the general public when not occupied with governmental service.² One of his weapons is mentioned in the will of Henry Bowcocke of Williamsburg. He left "1 bird piece made by Brush" to his son Henry, valued at two pounds, ten shillings.³ In 1723, however, Brush was "blown up and hurt" while firing off a cannon salute for the King's birthday. The injury was serious enough to warrant a petition to the House of Burgesses from Brush for "some Allowance," since apparently he could not carry on his trade.⁴ The petition was rejected and it is unknown whether he ever recovered his full health. Brush died three years later, in November or December of 1726. He was survived by a wife and four children.⁵

Brush's forge and workshop were most likely located on one of his two lots on the Palace Green (#165 and #166). A probate inventory made shortly after his death reveals that Brush was occupied in a wide range of activities. The presence of an anvil, a set of bellows, seven pair of smith's tongs, a pair of metal shears, 112 bushels of seal coal and a number of hammers, sledges, punches, and chisels indicate that he was involved with actually forging guns and gun parts instead of merely repairing or assembling them from pieces imported from England.⁶ The shop also held 23 pounds of steel and over 60 pounds of wrought iron which served as the raw materials for his work. Tools specific to gunsmithing listed in the inventory include six screw plates, six pairs of clamps, two tumbler tools, one "powder trier," a drawbore for boring holes in musket barrels, two pan borers and a "Cypharington hand vice."⁷ At the time of Brush's death, 128 the shop contained three gun barrels, two long shank bills, three rake heads, and a finished gun, all of which were probably either commissioned work or objects awaiting repair.⁸

Brush was also casting both brass and pewter objects at his shop. The inventory records twenty-nine melting pots, 172 pounds of "old brass," 45 pounds of copper, 35 pounds of pewter and 12 "new cast brass."⁹ The tools in Brush's shop also included four vices of varying sizes, two grindstones, a large wheel "for razor grinding," and at least 37 files for finishing forged or cast pieces and for sharpening swords and other bladed weapons.¹⁰ The space needed for operating some of these tools, particularly the drawbore mentioned above and the grinding stones, indicates that his workshop must have been fairly large.

In addition to forging the barrel and lock plate, Brush seems to have fashioned the wooden stocks for his guns. He owned a number of gougers, formers, and planes, as well as a spoke shave which he used in forming musket and pistol stocks. One finished wooden stock was found in his shop at the time of his death, perhaps awaiting assembly.¹¹

Brush's will and probate both indicate that he had neither slaves nor formal apprentices working for him, since they would have been listed as assets of the estate. He might have been aided in his work by his son, Anthony, or perhaps by his son-in-law, Thomas Barber. Neither of them carried on the gunsmithing trade after Brush's death, and Williamsburg was to go without a skilled gunsmith until the arrival of James Geddy in 1737.¹²

The Archaeological Record

The artifacts relating to John Brush's metalworking activities that were used in this study were excavated in -1987 and 1988 by Colonial Williamsburg archaeologists. Although other archaeology has been done on this site, suitable records were not kept. Artifactual evidence in the form of tools, gun parts, and metalworking by-products both confirms activities hinted at in the scant documentary record pertaining to Brush and reveals practices that are not mentioned.

The material culture associated with Brush's metalworking activities falls into five categories: tools associated with smithing, gun parts, gun accoutrements, non-gun forge work, and forging by-products. For the purposes of artifactual analysis, the objects examined were recovered from firmly dated stratigraphic contexts dating from 1720-1740 unless otherwise stated.

It is unfortunate that the shop itself has not been located. Features associated with Brush's shop either remain to be found or have been obliterated by earlier excavations. This makes definitive statements about shop layout impossible, but information can still be gleaned from the artifacts which have been recovered. The presence of coal, clinker, and slag in features on the property dating to Brush's period indicate that his shop was located here rather than elsewhere in Williamsburg.¹³ His workshop was perhaps dismantled by one of the property's later owners, spreading the associated artifacts across the site.

What is immediately striking when viewing the Brush gunsmithing assemblage is the relatively small size of the sample. Only a handful of gun parts were recovered when compared to the large assemblage associated with the James Geddy shop and foundry which opened by 1737 and was operated by Geddy and his sons into the 1750s.¹⁴ There are several explanations for the discrepancy in gun part representation, however. The Brush site potentially has a large 129 number of gun parts yet to be recovered if the forge is located, perhaps putting the Brush assemblage on a par with that of the Geddy site. Also, the scale of operation between the two craftsmen differed greatly; Brush worked alone or perhaps with his son or son-in-law, while Geddy employed himself, at least two of his four sons (William and David), a white indentured servant named William Beadle and probably a slave boy named Jack.¹⁵ The Geddy shop was also in operation for a longer period and saw a greater volume of business than Brush's operation. Geddy also benefitted from a predominantly civilian clientele while Brush's primary commitment was to the Virginia colonial government.

The best artifactual sources for reconstructing Brush's activities are gun parts, stock furniture, and accoutrements, since they represent the majority of non-architectural metal artifacts. As defined in this study, gun parts are those mechanisms which are necessary for the firing of pistols and muskets. These include barrels, lock plates and pans, cocks, frizzens and pan covers, springs, sears, tumblers, and triggers. Stock furniture are defined as gun elements not essential for firing, such as thumb and stock plates, fragmentary and complete trigger guards and terminals, thimbles and worms. Gun accoutrements consist of the flints and shot needed to complete the firing process.

Conclusion

John Brush, as one of the few gunsmiths in Virginia and the only gunsmith in Williamsburg at this time, was involved in 132 both repairing old guns and manufacturing new ones for private citizens, the colonial government and perhaps even the Indian Trading Company. He was converting older guns into flintlocks. He apparently cast shot for the militia and more decorative gun elements for privately-owned guns at his shop out of lead and brass. He was a consumer of French gunflints, which seem to dominate this period. His shop, located on this site, was fairly large despite the fact that he probably worked alone or with, at most, only one or two assistants. An accident in 1723 apparently resulted in a decline in his work due to injury, or perhaps to a shift towards working on privately-owned weapons.

Brush's tenure in Williamsburg, while not long, gives us a glimpse of a craftsman employed in a wide variety of activities. Although he was a gunsmithing specialist, Brush nevertheless produced the whole range of parts and accessories necessary to manufacture a musket or pistol. He dabbled in woodworking, casting, and cutlery in addition to his main occupation, forging.

Greater specialization would come later in the century with the advent of the Industrial Revolution, but during this period the lone craftsman was sovereign.

Endnotes

Laboratory Methods

The initial processing phase included sorting the faunal fragments into "identifiable" and "unidentifiable" categories. The unidentifiable bone-that which could not be taken to at least the taxonomic level of Order-was then further sorted into broad taxon groupings such as bird, fish, small mammal, medium mammal, and large mammal. Finally, within their taxon groupings, the bones were sorted into broad element categories such as long bones, teeth, ribs, and skull fragments. All of the unidentifiable bones were then counted, weighed, and examined for evidence of burning, butchering or other types of modification. This data was then entered into a custom-designed microcomputer program developed by Gregory Brown and Joanne Bowen for Colonial Williamsburg's Department of Archaeological Research.

Each of the identifiable bones was assigned a "unique bone number" which was affixed to the bone with an acid-free label along with the site and feature number. If the bones were too small to be labeled, they were placed into plastic bags along with an acid-free label. By working with a comparative skeletal collection created and maintained by Joanne Bowen of Colonial Williamsburg's Zooarchaeology Lab, the "identifiable" bone fragments were identified to the lowest taxonomic level possible. The taxon, bone element, symmetry (side), location, weight, fusion state, tooth type and wear, relative age, butchering techniques, and evidence of burning, weathering, and chewing were recorded and entered into the computer program. Once the data were entered, they were manipulated to provide calculations and percentages concerning bone size, bone frequency, and kill-off patterns of the faunal assemblage.

Once the identification was completed, the bones were laid out for minimum number of individuals (MNI) determination. MNI figures (see below) were calculated by pairing comparable rights and lefts, taking into account size, fusion state, tooth eruption, and general morphology. Before the bones were put away, osteological measurements of the major domesticates (pig, cow, and 136 sheep/goat) were made using the standards defined in von den Dreisch (1976).

Analytic Techniques

Zooarchaeologists have devised several methods of quantification to estimate relative dietary importance and to help adjust for differential preservation. These methods include the Number of Identified Specimens (NISP), Minimum Number of Individuals (MNI), minimum weight estimates, and biomass estimates. Individually, each method examines the faunal assemblage on a different level, but when used together, they help to establish a check-and-balance system to faunal interpretations of a site. Unfortunately, some zooarchaeologists have abandoned one or more methods used to measure the diversity of a faunal assemblage. By computing all four estimates, a comprehensive examination of the faunal assemblage is achieved, which has allowed the comparison of our data with the work of others, however limited these comparisons might be.

At the simplest level, the number of identified specimens (NISP) is used to calculate the relative abundance of any species within a faunal assemblage. After identification, all the bones within each species are added together to determine the frequency of fragments for each animal. This method is still often used, although it has several shortcomings, most notably its failure to account for element interdependence, differential preservation, variability in the identifiability of certain elements, and differences in collection techniques (Grayson 1984).

The most popular method for estimating category abundance is the statistic called Minimum Number of Individuals (MNI). While NISP attempts to calculate the maximum number of individuals on a site, MNI most often establishes the minimum number of animals by examining the most common element for each taxon. It provides a conservative approach to estimating the smallest number of animals that are represented in the recovered faunal assemblage. The MNI calculations can be made even more effective by carefully matching rights and lefts and by using age and sex indicators as well.

There are some other considerations such as the thoroughness of the analyst, the units of aggregation, and the sample size which can affect the interpretation of a faunal sample when using the MNI approach. Often, the least common species on a site will be overemphasized in small samples when using the MNI approach. This can provide a skewed picture of the relative dietary importance, as well as the fact that the MNI method regards large and small taxa has being of equal importance. For example, one pig and one cow are seen as equally important in dietary terms, despite the differences in pounds of meat. They provide an estimate of the relative importance of individuals, not meat, and they therefore do not necessarily reflect the dietary importance of different taxa (Grayson 1984).

The final technique which is quickly becoming a standard procedure in zooarchaeological analysis is known as the "biomass" or "skeletal mass allometry" method because it arrives at a percentage of meat weight based on the weight of the archaeological bone. Largely developed by Elizabeth Reitz of the University of Georgia and scholars at the University of Florida, this method is based on the basic formula for allometry, which assumes that any two dimensions of an animal grow at an exponential rather than a linear rate. Body size and body weight can then be determined from the size of a bone element, since a specific quantity of bone represents a predictable amount of tissue (Reitz and Cordier 1983; Reitz and 137 Scarry 1985). This estimate , therefore, provides a balance to the NISP and MNI methods. It successfully counters the problem of interdependence, since it accounts for the presence/absence of partial and complete skeletons. It does not rely on thoroughness or assemblage composition, and fragmentation is not a problem. It does, however, require that each bone (or set of bones) be weighed individually.

Many analysts have also used the distribution of particular elements to suggest important conclusions regarding taphonomy and/or butchering practices (e.g., Maltby 1979). Detailed studies of the location, orientation, and depth of butchering marks and carnivore or rodent chewing, beyond the scope of this analysis, are another method of investigating food preparation and disposal.

Animal husbandry is revealed by so-called "kill-off" patterns, based on epiphyseal fusion of (mostly) mammal long bones (Chaplin 1971; Payne 1973; Bowen 1994). Since the time of epiphyseal fusion is generally relatively constant within a species, an age distribution can be constructed for the identified animals (and by extension for the herds from which they came).

Environment is generally suggested by the diversity and relative abundance of certain wild taxa, particularly those with narrow ranges of ecological tolerance. In many cases seasonality can be revealed as well by looking at the presence and abundance of migratory species, such as waterfowl, as well as age patterns of domestic animals.

Excavation Methods

The bones upon which this analysis was performed came from soil screened through one-quarter-inch mesh, and it appears that even very small and fragile bone was collected. Quarter-inch screening is standard technique on sites dug by Colonial Williamsburg. It has been shown (Thomas 1969) that screening has an enormous positive influence on the recovery of bone and particularly in the recovery of smaller or more fragile species.

Description of Identified Taxa

At least eighteen taxa were identified in the assemblage. A brief description of each identified taxon is given below.

Relative Dietary Importance

Like virtually all assemblages from the eighteenth-century Chesapeake, the assemblage is predominated by cow (69.6% of the total biomass). Pig is next in line (11.6%), followed on a much smaller scale by sheep or goat (1.7%), domestic birds (0.4%), deer (0.3%), and fish (0.2%).

Although a relatively wide variety of taxa were recovered, including a few wild mammals and birds, turtles, and fish, the diet was overwhelming based on domestic mammals (84.2% of the total biomass). Even the large percentage of the biomass represented by domestic mammals undoubtedly understates the case; much, in fact probably just about all, of the 12.0% representing undifferentiated large and medium mammal must have come from cattle and swine, judging by the identifiable elements. Again, these large percentages are similar to those from other local sites, and certainly reflects a strong, consistent pattern that has its roots in husbandry practices, wildlife availability, cuisine preferences, and marketing.

Butchering Practices

No attempt was made to study changes in butchering techniques, although it was noted that a majority of the domestic mammal bones, particularly the long bones, had been hacked by an axe or cleaver, resulting in irregular fractures as well as shallow-to-deep V-shaped cuts. The most dramatic change in butchering technique-the advent of the mechanical band saw around 1850 or so-was not noted, suggesting that there was no serious contamination by later deposits.

Note: NISP= Number of identified specimens; MNI= Minimum number of individuals. "2/2" under MNI means 2 adult, 2 immature; "1" means 1 adult.

Those bones that were butchered most consisted of, not surprisingly, the major meat-bearing elements. Upper leg bones were often broken into several pieces; vertebrae were commonly split longitudinally through what would have been the midline of the body. Butchering marks were noted most often on cattle and pig bones. Little butchering was done on birds and fish; many were undoubtedly broken up by hand or thrown whole into the pot.

Element Distributions

By suggesting what parts of the animal were predominantly being eaten, element distributions can sometimes reveal patterns of use that are not shown by the relative proportions of biomass alone. As Table 2 shows, for example, more pig remains came from foot elements than would be expected (foot elements making up 37.3% of a normal pig skeleton, but comprising 50.9% of the Brush assemblage). The presence of large numbers of head and foot elements, relative to the "meatier" body elements, suggests that Brush was most likely keeping, and slaughtering, pigs on his property. The higher percentage of "body" elements for cow and sheep or goat, on the other hand, suggests that Brush was obtaining this meat from elsewhere-purchasing the meaty cuts,

Figure 1. Kill-off pattern for domestic swine (Sus scrofa).

Figure 2. Kill-off pattern for domestic cattle (Bos taurus)

Animal Husbandry

So-called "kill-off" patterns give another picture of animal husbandry. By charting the fusion status of the epiphyses, or ends of long bones, it is possible to determine some of the characteristics of the slaughtered population.

In this case it appears that the majority of the pigs were killed when they were three years or older (Fig. 1). This is an unusual distribution in the Chesapeake.

There were two peaks, on the other hand, of cattle slaughtering. About a third were killed at before 12 months; these calves would have been raised for veal. The remainder were killed after three years of age, presumably after attaining their maximum weight and being put to some other use, perhaps as dairy cows.

Conclusion

The bones from the Brush privy are typical in many ways of early eighteenth-century Virginia diet, as reconstructed from several years of zooarchaeological research. Cow and pig predominate, but lesser amounts of fish, turtle, wild game, and domestic birds are present. The occupants of this site were clearly heavily dependant on domestic livestock, to the extent that they failed to take advantage of an abundance of locally-plentiful fish, birds, and wild mammals.

References

• Behler, John, and F. Wayne King (1979). Audubon Society Field Guide to North American Reptiles and Amphibians. Alfred A. Knopf, New York.
• Bent, Arthur Cleveland (1963). Life Histories of North American Gallinaceous Birds. Reprint of 1932 edition. Dover Publications, Inc., New York.
• Bidwell, Percy W., and John L. Falconer (1925). History of Agriculture in the United States 1620-1860. Reprinted 1941, Peter Smith, New York.
• Bowen, Joanne (1989). A Comparative Analysis of the New England and Chesapeake Herding Systems: The Relative Dietary Importance of Beef and Dairy Products. Paper presented at the 22nd annual meeting 145 of the Society for Historical Archaeology, Baltimore.
• Bowen, Joanne (1994). "A Comparative Analysis of the New England and Chesapeake Herding Systems." In The Historic Chesapeake: Archaeological Contributions, edited by Paul A. Shackel and Barbara J. Little, pp. 155-167. Smithsonian Institution Museum Press, Washington, D.C.
• Chaplin, Raymond (1971). The Study of Animal Bones from Archaeological Sites. Seminar Press, London.
• Chapman, Joseph, J. Gregory Hockman, and William Edwards (1982). "Cottontails: Sylvilagus floridanus and allies." In Wild Mammals of North America: Biology, Management, and Economics, edited by Joseph Chapman and George Feldhamer, pp. 83-123. The John Hopkins University Press, Baltimore.
• Conant, R. (1975). A Field Guide to Reptiles and Amphibians of Eastern and Central North America. Houghton Mifflin Company, Boston.
• Ernst, Carl H., and Roger W. Barbour (1972). Turtles of the United States. The University Press of Kentucky, Lexington.
• Flyger, Van, and J. Edward Gates (1982). "Fox and Gray Squirrels (Sciurus niger, S. carolinensis, and Allies)." In Wild Mammals of North America: Biology, Management, and Economics, edited by Joseph Chapman and George Feldhamer, pp. 209-229. The Johns Hopkins University Press, Baltimore.
• Gray, Lewis C. (1958). History of Agriculture in the Southern United States to 1860. Carnegie Institution of Washington Publication No. 430. Carnegie Institution, Washington, DC.
• Grayson, Donald K. (1984). Quantitative Zooarchaeology: Topics in the Analysis of Archaeological Faunas. Academic Press, Orlando.
• Herald, Earl S., editor (n.d.). Fishes of North America. Doubleday & Company, New York.
• Hesselton, W., and R. Hesselton (1982). "White-tail Deer, Odocoileus virginianus." In Wild Mammals of North America, edited by J. Chapman and G. Feldhamer, pp. 878-901. The Johns Hopkins University Press, Baltimore.
• Jackson, William B. (1982). "Norway Rat and Allies (Rattus norvegicus and Allies)." In Wild Mammals of North America, edited by Joseph Chapman and George Feldhamer, pp. 1077-1088. The Johns Hopkins University Press, Baltimore.
• Lippson, Alice Jane, and Robert L. Lippson (1984). Life in the Chesapeake Bay. The Johns Hopkins University Press, Baltimore.
• McClane, A.J., editor (1965). McClane's Field Guide to Freshwater Fishes of North America. Holt, Reinhart and Winston, New York.
• McVeagh, Lincoln, editor (1924). The Journal of Nicholas Cresswell, 1774-1777. The Dial Press, New York.
• Miller, Henry M. (1984). Colonization and Subsistence Change on the 17th Century Chesapeake Frontier. Ph.D. dissertation, Michigan State University, Lansing. University Microfilms, Ann Arbor.
• Noìl Hume, Audrey (1978). Food. Colonial Williamsburg Archaeological Series No. 9. The Colonial Williamsburg Foundation, Williamsburg.
• Payne, Sebastian (1973). "Kill-Off Patterns in Sheep and Goats: The Mandibles from Asvan Kale." Anatolean Studies 23:281-303.
• Powell, Richard (1990). Personal communication with Joanne Bowen. Williamsburg, Virginia.
• Reitz, Elizabeth J. (1979). Spanish and British Subsistence Strategies at St. Augustine, Florida and Frederica, Georgia Between 1565 and 1783. Ph.D. dissertation, University of Florida, Gainesville. University Microfilms, Ann Arbor.
• Reitz, Elizabeth J., and Dan Cordier (1983). Use of Allometry in Zooarchaeological Analysis. In Animals and Archaeology: 2. Shell Middens, Fishes and Birds, edited by Caroline Grigson and Julia Clutton-Brock, pp. 237-252. B.A.R. International Series 183, London.
• Reitz, Elizabeth J., and C. Margaret Scarry (1985). Reconstructing Historic Subsistence with an 146 Example from Sixteenth-Century Spanish Florida. Special Publication Series, Society for Historical Archaeology No. 3. Society for Historical Archaeology, Ann Arbor.
• Silver, I.A. (1969). "The Ageing of Domestic Animals." In Science in Archaeology: A Survey of Progress and Research, edited by Don Brothwell and Eric Higgs, pp. 283-302. Praeger Publishers, New York.
• Thomas, David Hurst (1969). Great Basin Hunting Patterns: A Quantitative Method for Treating Faunal Remains. American Antiquity 34(4):393-401.

Appendix 3, Table 3.
Kill-off Data
Sus scrofa (Domestic Pig)
N=108

Appendix 3, Table 4.
Kill-off Data
Bos taurus (Domestic Cow)
N=58

References Cited

• Beaudry, Mary, and Stephen Mrozowski (editors) (1989). Interdisciplinary Investigations of the Boott Mills Lowell, Massachusetts. Volume III: The Boardinghouse System as a Way of Life. Report prepared for the National Park Service by the Center for Archaeological Studies, Boston University.
• Carr, Lois G., and Lorena S. Walsh (1994). Changing Lifestyles and Consumer Behavior in the Colonial Chesapeake. In Of Consuming Interests: The Style of Life in the Eighteenth Century, edited by Cary Carson, Ronald Hoffman, and Peter J. Albert, pp. 59-166. University Press of Virginia, Charlottesville.
• Hess, Karen (editor) (1981). Martha Washington's Booke of Cookery. Columbia University Press, New York.
• Hess, Karen (editor) (1984). Virginia Housewife, by Mary Randolph. University of South Carolina Press, Columbia. Facsimile of 1824 edition.
• Langley, Batty (1728). New Principles of Gardening. A. Betteswoth and J. Batley, London.
• Markham, Gervase (1615). The English Housewife. Edited by Michael Best, 1986. McGill-Queen's University Press, Kingston.
• Mrozowski, Stephen A. (1991). A Preliminary Report of the Archaeobotanical Analysis of Context 29G-The Brush-Everard Site, Williamsburg, Virginia. Ms. on file, Department of Archaeological Research, Colonial Williamsburg Foundation, Williamsburg, Virginia.
• Pearsall, Deborah M. (1989). Paleoethnobotany; A Handbook of Procedures. Academic Press, New York.
• Randolph, John, Jr. (1826). A Treatise on Gardening by a Citizen of Virginia. Edited by M. F. Warner. Reprinted from The American Gardener of John Gardiner and David Hepburn. 3rd edition. Appeals Press, Richmond, 1924.
• Reinhard, Karl (1989). Analysis of Latrine Soils from the Brush-Everard Site, Colonial Williamsburg, Virginia. Ms. on file, Department of Archaeological Research, Colonial Williamsburg Foundation, Williamsburg, Virginia.
• Reinhard, Karl (1991). Parasitological Analysis of the Brush-Everard Site, Williamsburg, Virginia. Ms. on file, Department of Archaeological Research, Colonial Williamsburg Foundation, Williamsburg, Virginia.
• Reinhard, K.J., S.A. Mrozowski, and K.A. Orloski (1986). Privies, pollen, parasites and seeds, a biological nexus in historic archaeology. MASCA Journal 4:31-36.
• Webb-Prentis Papers. Manuscripts on file, Colonial Williamsburg Foundation Library, Williamsburg.

Materials and Methods

The soil samples were processed for the recovery of both pollen and parasite eggs. The process is designed to dissolve and/or extract different soil components to leave specific organic debris containing parasite eggs and pollen. Two subsamples were extracted from each soil sample sent to the laboratory, one for parasite analysis and one for pollen analysis. The pollen sample and parasite sample are processed in the same way except that the pollen samples are submitted to additional chemical treatments that destroy parasite eggs.

Thirty milliliters of soil were measured for each subsample. A Lycopodium spore tablet was added to each subsample. Each spore tablet contains 11,300 plus or minus 200 spores. The addition of a known number of spores to the subsamples enables accurate measurement of the number of parasite eggs or pollen grains per milliliter of soil by calculating a ratio of eggs or pollen to Lycopodium spores.

The subsamples were treated first in 30% hydrochloric acid in 300 milliliter beakers. The acid dissolves calcium carbonate that holds microscopic particles in the soil matrix. Acid was added to the soil samples until reaction between the acid and carbonates ceased. Then distilled water was added to the beakers holding the soil-acid mixture.

Once the soil was disaggregated by hydrochloric acid, they were sedimented and screened to remove large, heavy components. The soil mixture was rigorously swirled until the soil was in suspension. The beaker was then set aside for 30 seconds to allow the heavy fraction to settle. The supernatant was then poured through a 0.25 millimeter mesh screen into a 500 milliliter beaker. This process was repeated twice, at which point the supernatant was nearly 154 clear. The material resting on top of the screen was dried on blotter paper and then examined for macroscopic remains. The heavy sand sediment was discarded.

The microscopic remains in the 500 milliliter beakers were concentrated by centrifugation. The concentrated remains were then washed three times in distilled water to remove traces of hydrochloric acid that would otherwise react with chemicals in subsequent stages of processing.

After the microscopic remains were washed and again concentrated by centrifugation, they were treated with 72% hydrofluoric acid. This process dissolves fine silicates. The concentrated remains were transferred to 700 milliliter plastic beakers and about 50 milliliters of acid were added to the soil. After stirring, the soil-acid mixture was set aside for 24 hours to allow for completion of the reaction. After 24 hours, the remaining sediments were concentrated by centrifugation. The sediments were then washed with distilled water to remove residual hydrofluoric acid that would otherwise pose a health hazard to the analyst.

After the water washes, distilled water was added to the microscopic sediments in 50 milliliter centrifuge tubes The tubes were then placed in a sonicator and sonicated for 4 minutes. This treatment loosens fine organic debris and separates the microscopic particles. After sonication, the microscopic remains were transferred to 12 milliliter glass centrifuge tubes. After the microscopic remains were concentrated by centrifugation and the supernatant poured off, a zinc bromide heavy density mixture (specific gravity 2.0) was added to the tubes. The sediment was then mixed into the zinc bromide and the tubes were spun in a clinical centrifuge at 1,500 r.p.m. for 15 minutes. This process resulted in the separation of light organic remains, including pollen grains and parasite eggs, from heavier organic detritus. The light remains floated to the surface of the heavy density mixture and were easily removed. The heavy detritus sank to the bottom of the tubes.

After the zinc bromide heavy density treatment, one subsample of each soil sample was transferred to glass vials in glycerol and examined for parasite eggs. The other subsamples were processed further to extract pollen by acetolysis. The acetolysis step destroys parasite eggs and consequently, the parasite subsample underwent no further processing.

The pollen subsamples were washed twice in glacial acetic acid. Then an acetolysis mixture (9 parts acetic anhydride to one part sulfuric acid) was added to the tubes which were then heated for 20 minutes. The acetolysis treatment dissolves several organic compounds, the most important of which are cellulose and chitin. After the acetolysis treatment, the soils were washed once with glacial acetic acid and then with distilled water until the supernatant was clear.

The microscopic remains were then treated for 30 seconds in 5% potassium hydroxide to dissolve humic compounds. After several water washes the supernatant was clear and the microscopic remains were transferred into vials with glycerol.

Microscopic examination of both the parasite and pollen samples was accomplished by placing a drop of glycerol with suspended microscopic remains onto a microscope slide. A coverslip was placed over the drop and sealed with commercial nail polish. After the polish dried, the slides were examined with a binocular compound microscope. The pollen preparation was examined at 400 power and the parasite sample was examined at 200 power. The differences in magnification are due to the general larger size of parasite eggs which can easily be discerned at lower magnification in comparison to smaller pollen.

The first goal in examining the pollen preparations was to determine whether or not the samples contained enough pollen for a full 200 grain count. It has been found by researchers in the Palynology Laboratory, Texas A&M University that less than 1,000 pollen grains per milliliter of soil sample is insufficient for a full 200 grain count. This is due to two factors. First, samples that contain less than l,000 grains are usually subject to water percolation or decomposition. In such cases there is differential preservation of pollen types. Thus, the resulting pollen count presents a skewed representation of the pollen types originally present in the soil. Secondarily, it has been found that counting pollen from samples containing less than 1,000 grains is extremely time consuming. Therefore, the analyst spends an uneconomical amount of time to obtain counts that have little scientific validity.

The pollen and eggs were counted as were Lycopodium spores. A minimum of two hundred pollen grains were counted from samples that contained sufficient pollen. The 200 grain count has been found to be statistically reliable at the 95% confidence interval (Barkley 1934) The number of pollen grains per milliliter of soil was calculated on the basis of the ration of pollen grains to the known number of Lycopodium spores. The same process was applied in the parasite egg analysis.

Results

The dried macroscopic remains from sample I contained sand and fragments of what I believe to be anthracitic coal or coke derived from the burning of coal. Sample 2 contained fine sand and many Rubus seeds. Rubus seeds are commonly found in Colonial and post-Colonial latrines in Newport, Rhode Island, and Greenwich Village, New York. Their presence in the Brush-Everard deposits is, therefore, suggestive of privy deposits.

No parasite eggs were found in sample one. However, sample 2 contained 1,200 eggs per gram (Table 1). Of these, 83% are identical to the eggs of the human whipworm, Trichuris trichiura. The remainder are similar to the eggs of the giant intestinal roundworm, Ascaris lumbricoides.

The pollen concentration values differed between the two soil samples. In sample 1, 129 pollen grains per milliliter of soil were present. In sample 2, 20,100 grains per gram were present. Thus, there was insufficient pollen in sample I to allow for a complete 200 grain count. However, there was an abundance of pollen in sample 2.

The pollen spectrum of sample 2 was dominated by what I believe to be a species in the Caparidaceae (Caper family). Of 1,011 pollen grains counted, 869 were of this type (Tables 2 and 3). This identification is not, 156

Other dietary types are present in the deposits. Most importantly, large grains of a cultivated grass are present. The grains are consistent with the morphology of maize. Three of 12 cultivate grass grains are broken. Such breakage is indicative of grinding. Thus, these pollen indicate 157

Many environmental pollen types were present in the privy. These include arboreal plants (Salix, Quercus, Nyssa, Populus, Pinus, Carya, Liquidambar). Most of these are derived from trees growing in the Williamsburg area. Pinus, however, produces exceptionally buoyant pollen grains that are carried many miles from their source. Consequently, Pinus pollen is not necessarily derived from trees in the Williamsburg area. Nonarboreal plants are also represented in the pollen spectrum. These include Cheno Ams (Chenopod and Amaranth families), types in the Asteraceae (composite family), Poaceae (grass family) and Berberis (barberry). These types originated from plants growing in the vicinity of Williamsburg.

Discussion and Summary

The parasite eggs include species that are transferred by contact with human excreta. It is clear that sanitation practices in Williamsburg were not sophisticated enough to prevent fecal borne disease. The presence of the eggs of intestinal worms in the privy is proxy evidence that the inhabitants of Williamsburg were at risk of protozoal and bacterial infection with species that are transferred by direct fecal contact. It is therefore probable that infections with Giardia sp., Entamoeba hvstolytica, Iodamoeba sp. and other protozoa were common and that 158 children especially were susceptible to diarrheal disease from these species as well as bacterial pathogens.

The pollen evidence documents the consumption of flowers such as capers which may have been eaten as a spice and were possibly introduced from Europe. More evaluation of the pollen is necessary to support this hypothesis. Cultivated grains, possibly maize, also constituted part of the diet.

The environmental information indicates that a variety of trees grew in the Williamsburg area. These included black gum, sweet gum, oak, willow, cottonwood and pecan. However, non- arboreal plants are more heavily represented in the pollen record. Most important are pollen grains derived from disturbance types such as Chenopodiaceae, Amaranthaceae, low spine Asteraceae (ragweed type), Artemisia, and Poaceae (grass family). The dominance of these types of pollen indicate that environment in the area of the Brush-Everard site was disturbed by gardening, plowing, or other intensive human activity.

The find of parasite eggs, seeds and high quantities of pollen is clear evidence that the soils submitted for analysis are derived from fecal deposits. This confirms the hypothesis that the feature under study did serve as a latrine as well as a kitchen midden.

Literature Cited

• Barkley, F.A. (1934). The statistical theory of pollen analysis. Ecology 15:283-289.
• Herrmann, B. (1986). Parasitologische Untersuchung mittelalterlicher Kloaken. In B. Herrmann (ed.), Mensch und Umwelt im Mittelalter. Stuttgart: Deutsche Verlags-Anstalt, pp. 160-169.
• Jones, A.K.G. (1985). Trichurid ova in archaeological deposits: their value as indicators of ancient feces. In N.J.R. Fieller, D.D. Gilbertson, and N.G.A. Ralph (eds.), Paleobiological Investigations: Research Design, methods and data analysis. Symposia of the Association for Environmental Archaeology No. 5b, BAR International Series 266, pp. 105-119.
• Jones, A.K.G., A.R. Hutchinson, and C. Nicholson (1988). The worms of Roman horses and other finds of intestinal parasite eggs from unpromising deposits. Antiquity 62:225-229.
• Reinhard, K.J., U.E. Confalonieri, B. Herrmann, L.F. Ferreira, and A.J.G. Araujo (1988). Recovery of parasite eggs from coprolites and latrines: aspect of paleoparasitological technique. Homo 37:217-239.
• Reinhard, K.J., S.A. Mrozowski, and K.A. Orloski (1986). Privies, pollen, parasites and seeds, a biological nexus in historic archaeology. MASCA Journal 4:31-36.

Materials and Methods

Eleven samples were analyzed, seven from the site's privy, from the ravine, one from the refuse layer, and one other control sample.

The technique employed here is very useful in quantifying small numbers of eggs in soil deposits and therefore in tracing areas of archaeological sites that were potentially infective. Thirty milliliters of soil were extracted from each soil sample. To each 30 ml sample, a Lycopodium spore tablet contain 11,400 plus or minus 400 spores was added. The samples were then treated with hydrochloric acid to dissolve carbonates and liberate microscopic particles. The samples were then screened through a mesh of approximately 300 micrometers. The fine microscopic particles that passed through the screen were then concentrated by centrifugation. The sediments were then floated in a zinc bromide heavy density solution of 2.0 specific gravity. The supernatant containing light, microscopic remains was poured off and concentrated through centrifugation. The microscopic remains were partially dehydrated in 50% ETOH and then transferred to 2 dram vials in glycerol.

To make microscope preparations, drops of the glycerol containing sediments were placed on microscope slides and covered with a cover glass. Several preparations of each sample were made and examined under 200 power. For those samples that contained parasite eggs, the Lvcopodium spores were counted and a ratio of spores to parasite eggs was calculated. Based on the known number of spores per ml of sediment, the number of eggs per ml were calculated.

Results

Only two of the samples, both from the latrine, contained parasite eggs. These were samples 29G-637 and 29G-640. Sample 29G-640 contained 171 eggs per ml of soil. Sample 29G-637 contained 262 eggs per gram. Of these two samples, only eggs of A. lumbricoides were found.

Discussion

The lack of parasite eggs in most soil samples indicates that the site was not extensively saturated with parasite eggs. It would be interesting to examine house floors and other special use areas to ascertain whether parasite eggs were concentrated in areas where human activity was most intense.

The egg concentration values for the samples that were parasite positive are both well below those defined by Jones (1985) as being typical of fecal soils in an urban setting. Jones states that 400 eggs per gram are typical of the urban environment. However, the European work only addresses parasite ecology in the urban setting.

In the less urban conditions of early colonial America, especially on farmsteads, one would not anticipate finding many, if any, parasite eggs throughout a site. The parasite ecology of early American rural colonists must have been defined not so much by population crowding but rather by specific and possibly idiosyncratic behaviors that maintained parasite life cycles in circumscribed areas. Thus, the evaluation of rural Colonial parasite ecology must focus on sampling areas that can clearly be associated with human habitation and activity.

In contrast to the latrine sample analyzed previously (Reinhard 1989) which contained an abundance of parasite eggs of two species (1,200 eggs per gram), the samples described herein contain very few eggs of only one species. In the sample examined previously, the eggs of T. trichiura dominated the samples. The difference between the samples is suggestive of the effects of differential preservation on parasite eggs. As noted previously in Europe and North America (Reinhard et al. 1986), T. trichiura eggs seem to decompose faster than those of A. lumhricoides. The difference in the egg counts could be due to either differential preservation of to different parasite spectra of the household at different period in time. The low egg counts in this study make me suspect that differential preservation is a more likely cause. Thus, in the interpretation of parasite ecology from historic sites, it appears prudent to evaluate the data with a degree of skepticism and in context of differential preservation.

It is of interest that A. lumbricoides and T. trichiura eggs dominate the samples from the Brush-Everard Site. Ascarids, and trichurids, in the modern world are the most common human parasites and can persist in a variety of environments due to the extracorporal egg stage which is remarkable resistant do the environment.

This suggests that only the most infective fecal borne parasites could sustain infections at this site and that its parasite ecology was limited to only a few, fecal horne species.

References Cited

• Herrmann, B. (1986). Parasitologische Untersuchung mittelalterlicher Kloaken. In B. Herrmann (ed.), Mensch und Umwelt im Mittelalter. Stuttgart: Deutsche Verlags-Anstalt, pp. 160-169.
• Herrmann, B. (1987). Parasitologisch-Epidemiologische Auswertungen mittelalterliche Kloaken. Zeitschrift für Archaeologie des Mittelalters. K÷ln: Rheinland-Verlag, pp. 131-161. 161
• Herrmann, B. and U. Schultz (1986). Parasitologische Untersuchungen eines Spätmittelalterlich-frühneuzeitlichen Kloakeninhaltes aus der Lubecker Fronerei. Lubecker Schriften zur Archaologie und Kulturgeschichte 12:167-172.
• Jones, A.K.G. (1985). Trichurid ova in archaeological deposits: their value as indicators of ancient feces. In N.J.R. Fieller, D.D. Gilbertson, and N.G.A. Ralph (eds.), Paleobiological Investigations: Research Design, methods and data analysis. Symposium of the Association for Environmental Archaeology No. 5b, BAR International Series 266, pp. 105-119.
• Reinhard K.J., U.E. Confalonieri, B. Herrmann, L.F. Ferreira, and A.J.G. Araijo (1986). Recovery of parasite eggs from coprolites and latrines: aspect of paleoparasitological technique. Homo 37:217--239.
• Reinhard, K.J. (1991). Parasitology as an interpretive tool in Archaeology. American Antiquity 57(2):231-245.